You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
rocksdb/db/compaction/compaction_job.cc

2045 lines
79 KiB

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include "db/compaction/compaction_job.h"
#include <algorithm>
#include <cinttypes>
#include <memory>
#include <optional>
#include <set>
#include <utility>
#include <vector>
#include "db/blob/blob_counting_iterator.h"
#include "db/blob/blob_file_addition.h"
#include "db/blob/blob_file_builder.h"
#include "db/builder.h"
#include "db/compaction/clipping_iterator.h"
#include "db/compaction/compaction_state.h"
#include "db/db_impl/db_impl.h"
#include "db/dbformat.h"
#include "db/error_handler.h"
#include "db/event_helpers.h"
#include "db/history_trimming_iterator.h"
#include "db/log_writer.h"
#include "db/merge_helper.h"
#include "db/range_del_aggregator.h"
#include "db/version_edit.h"
#include "db/version_set.h"
#include "file/filename.h"
#include "file/read_write_util.h"
#include "file/sst_file_manager_impl.h"
#include "file/writable_file_writer.h"
#include "logging/log_buffer.h"
#include "logging/logging.h"
#include "monitoring/iostats_context_imp.h"
#include "monitoring/thread_status_util.h"
#include "options/configurable_helper.h"
#include "options/options_helper.h"
#include "port/port.h"
#include "rocksdb/db.h"
#include "rocksdb/env.h"
#include "rocksdb/options.h"
#include "rocksdb/statistics.h"
#include "rocksdb/status.h"
#include "rocksdb/table.h"
#include "rocksdb/utilities/options_type.h"
#include "table/merging_iterator.h"
#include "table/table_builder.h"
#include "table/unique_id_impl.h"
#include "test_util/sync_point.h"
#include "util/stop_watch.h"
namespace ROCKSDB_NAMESPACE {
const char* GetCompactionReasonString(CompactionReason compaction_reason) {
switch (compaction_reason) {
case CompactionReason::kUnknown:
return "Unknown";
case CompactionReason::kLevelL0FilesNum:
return "LevelL0FilesNum";
case CompactionReason::kLevelMaxLevelSize:
return "LevelMaxLevelSize";
case CompactionReason::kUniversalSizeAmplification:
return "UniversalSizeAmplification";
case CompactionReason::kUniversalSizeRatio:
return "UniversalSizeRatio";
case CompactionReason::kUniversalSortedRunNum:
return "UniversalSortedRunNum";
case CompactionReason::kFIFOMaxSize:
return "FIFOMaxSize";
case CompactionReason::kFIFOReduceNumFiles:
return "FIFOReduceNumFiles";
case CompactionReason::kFIFOTtl:
return "FIFOTtl";
case CompactionReason::kManualCompaction:
return "ManualCompaction";
case CompactionReason::kFilesMarkedForCompaction:
return "FilesMarkedForCompaction";
case CompactionReason::kBottommostFiles:
return "BottommostFiles";
case CompactionReason::kTtl:
return "Ttl";
case CompactionReason::kFlush:
return "Flush";
case CompactionReason::kExternalSstIngestion:
return "ExternalSstIngestion";
Periodic Compactions (#5166) Summary: Introducing Periodic Compactions. This feature allows all the files in a CF to be periodically compacted. It could help in catching any corruptions that could creep into the DB proactively as every file is constantly getting re-compacted. And also, of course, it helps to cleanup data older than certain threshold. - Introduced a new option `periodic_compaction_time` to control how long a file can live without being compacted in a CF. - This works across all levels. - The files are put in the same level after going through the compaction. (Related files in the same level are picked up as `ExpandInputstoCleanCut` is used). - Compaction filters, if any, are invoked as usual. - A new table property, `file_creation_time`, is introduced to implement this feature. This property is set to the time at which the SST file was created (and that time is given by the underlying Env/OS). This feature can be enabled on its own, or in conjunction with `ttl`. It is possible to set a different time threshold for the bottom level when used in conjunction with ttl. Since `ttl` works only on 0 to last but one levels, you could set `ttl` to, say, 1 day, and `periodic_compaction_time` to, say, 7 days. Since `ttl < periodic_compaction_time` all files in last but one levels keep getting picked up based on ttl, and almost never based on periodic_compaction_time. The files in the bottom level get picked up for compaction based on `periodic_compaction_time`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5166 Differential Revision: D14884441 Pulled By: sagar0 fbshipit-source-id: 408426cbacb409c06386a98632dcf90bfa1bda47
6 years ago
case CompactionReason::kPeriodicCompaction:
return "PeriodicCompaction";
case CompactionReason::kChangeTemperature:
return "ChangeTemperature";
Make it possible to force the garbage collection of the oldest blob files (#8994) Summary: The current BlobDB garbage collection logic works by relocating the valid blobs from the oldest blob files as they are encountered during compaction, and cleaning up blob files once they contain nothing but garbage. However, with sufficiently skewed workloads, it is theoretically possible to end up in a situation when few or no compactions get scheduled for the SST files that contain references to the oldest blob files, which can lead to increased space amp due to the lack of GC. In order to efficiently handle such workloads, the patch adds a new BlobDB configuration option called `blob_garbage_collection_force_threshold`, which signals to BlobDB to schedule targeted compactions for the SST files that keep alive the oldest batch of blob files if the overall ratio of garbage in the given blob files meets the threshold *and* all the given blob files are eligible for GC based on `blob_garbage_collection_age_cutoff`. (For example, if the new option is set to 0.9, targeted compactions will get scheduled if the sum of garbage bytes meets or exceeds 90% of the sum of total bytes in the oldest blob files, assuming all affected blob files are below the age-based cutoff.) The net result of these targeted compactions is that the valid blobs in the oldest blob files are relocated and the oldest blob files themselves cleaned up (since *all* SST files that rely on them get compacted away). These targeted compactions are similar to periodic compactions in the sense that they force certain SST files that otherwise would not get picked up to undergo compaction and also in the sense that instead of merging files from multiple levels, they target a single file. (Note: such compactions might still include neighboring files from the same level due to the need of having a "clean cut" boundary but they never include any files from any other level.) This functionality is currently only supported with the leveled compaction style and is inactive by default (since the default value is set to 1.0, i.e. 100%). Pull Request resolved: https://github.com/facebook/rocksdb/pull/8994 Test Plan: Ran `make check` and tested using `db_bench` and the stress/crash tests. Reviewed By: riversand963 Differential Revision: D31489850 Pulled By: ltamasi fbshipit-source-id: 44057d511726a0e2a03c5d9313d7511b3f0c4eab
3 years ago
case CompactionReason::kForcedBlobGC:
return "ForcedBlobGC";
case CompactionReason::kRoundRobinTtl:
return "RoundRobinTtl";
Add missing range conflict check between file ingestion and RefitLevel() (#10988) Summary: **Context:** File ingestion never checks whether the key range it acts on overlaps with an ongoing RefitLevel() (used in `CompactRange()` with `change_level=true`). That's because RefitLevel() doesn't register and make its key range known to file ingestion. Though it checks overlapping with other compactions by https://github.com/facebook/rocksdb/blob/7.8.fb/db/external_sst_file_ingestion_job.cc#L998. RefitLevel() (used in `CompactRange()` with `change_level=true`) doesn't check whether the key range it acts on overlaps with an ongoing file ingestion. That's because file ingestion does not register and make its key range known to other compactions. - Note that non-refitlevel-compaction (e.g, manual compaction w/o RefitLevel() or general compaction) also does not check key range overlap with ongoing file ingestion for the same reason. - But it's fine. Credited to cbi42's discovery, `WaitForIngestFile` was called by background and foreground compactions. They were introduced in https://github.com/facebook/rocksdb/commit/0f88160f67d36ea30e3aca3a3cef924c3a009be6, https://github.com/facebook/rocksdb/commit/5c64fb67d2fc198f1a73ff3ae543749a6a41f513 and https://github.com/facebook/rocksdb/commit/87dfc1d23e0e16ff73e15f63c6fa0fb3b3fc8c8c. - Regardless, this PR registers file ingestion like a compaction is a general approach that will also add range conflict check between file ingestion and non-refitlevel-compaction, though it has not been the issue motivated this PR. Above are bugs resulting in two bad consequences: - If file ingestion and RefitLevel() creates files in the same level, then range-overlapped files will be created at that level and caught as corruption by `force_consistency_checks=true` - If file ingestion and RefitLevel() creates file in different levels, then with one further compaction on the ingested file, it can result in two same keys both with seqno 0 in two different levels. Then with iterator's [optimization](https://github.com/facebook/rocksdb/blame/c62f3221698fd273b673d4f7e54eabb8329a4369/db/db_iter.cc#L342-L343) that assumes no two same keys both with seqno 0, it will either break this assertion in debug build or, even worst, return value of this same key for the key after it, which is the wrong value to return, in release build. Therefore we decide to introduce range conflict check for file ingestion and RefitLevel() inspired from the existing range conflict check among compactions. **Summary:** - Treat file ingestion job and RefitLevel() as `Compaction` of new compaction reasons: `CompactionReason::kExternalSstIngestion` and `CompactionReason::kRefitLevel` and register/unregister them. File ingestion is treated as compaction from L0 to different levels and RefitLevel() as compaction from source level to target level. - Check for `RangeOverlapWithCompaction` with other ongoing compactions, `RegisterCompaction()` on this "compaction" before changing the LSM state in `VersionStorageInfo`, and `UnregisterCompaction()` after changing. - Replace scattered fixes (https://github.com/facebook/rocksdb/commit/0f88160f67d36ea30e3aca3a3cef924c3a009be6, https://github.com/facebook/rocksdb/commit/5c64fb67d2fc198f1a73ff3ae543749a6a41f513 and https://github.com/facebook/rocksdb/commit/87dfc1d23e0e16ff73e15f63c6fa0fb3b3fc8c8c.) that prevents overlapping between file ingestion and non-refit-level compaction with this fix cuz those practices are easy to overlook. - Misc: logic cleanup, see PR comments Pull Request resolved: https://github.com/facebook/rocksdb/pull/10988 Test Plan: - New unit test `DBCompactionTestWithOngoingFileIngestionParam*` that failed pre-fix and passed afterwards. - Made compatible with existing tests, see PR comments - make check - [Ongoing] Stress test rehearsal with normal value and aggressive CI value https://github.com/facebook/rocksdb/pull/10761 Reviewed By: cbi42 Differential Revision: D41535685 Pulled By: hx235 fbshipit-source-id: 549833a577ba1496d20a870583d4caa737da1258
2 years ago
case CompactionReason::kRefitLevel:
return "RefitLevel";
case CompactionReason::kNumOfReasons:
// fall through
default:
assert(false);
return "Invalid";
}
}
const char* GetCompactionPenultimateOutputRangeTypeString(
Compaction::PenultimateOutputRangeType range_type) {
switch (range_type) {
case Compaction::PenultimateOutputRangeType::kNotSupported:
return "NotSupported";
case Compaction::PenultimateOutputRangeType::kFullRange:
return "FullRange";
case Compaction::PenultimateOutputRangeType::kNonLastRange:
return "NonLastRange";
case Compaction::PenultimateOutputRangeType::kDisabled:
return "Disabled";
default:
assert(false);
return "Invalid";
}
}
CompactionJob::CompactionJob(
int job_id, Compaction* compaction, const ImmutableDBOptions& db_options,
const MutableDBOptions& mutable_db_options, const FileOptions& file_options,
VersionSet* versions, const std::atomic<bool>* shutting_down,
LogBuffer* log_buffer, FSDirectory* db_directory,
FSDirectory* output_directory, FSDirectory* blob_output_directory,
Statistics* stats, InstrumentedMutex* db_mutex,
ErrorHandler* db_error_handler,
std::vector<SequenceNumber> existing_snapshots,
SequenceNumber earliest_write_conflict_snapshot,
CompactionIterator sees consistent view of which keys are committed (#9830) Summary: **This PR does not affect the functionality of `DB` and write-committed transactions.** `CompactionIterator` uses `KeyCommitted(seq)` to determine if a key in the database is committed. As the name 'write-committed' implies, if write-committed policy is used, a key exists in the database only if it is committed. In fact, the implementation of `KeyCommitted()` is as follows: ``` inline bool KeyCommitted(SequenceNumber seq) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(seq, kMaxSequence) == SnapshotCheckerResult::kInSnapshot; } ``` With that being said, we focus on write-prepared/write-unprepared transactions. A few notes: - A key can exist in the db even if it's uncommitted. Therefore, we rely on `snapshot_checker_` to determine data visibility. We also require that all writes go through transaction API instead of the raw `WriteBatch` + `Write`, thus at most one uncommitted version of one user key can exist in the database. - `CompactionIterator` outputs a key as long as the key is uncommitted. Due to the above reasons, it is possible that `CompactionIterator` decides to output an uncommitted key without doing further checks on the key (`NextFromInput()`). By the time the key is being prepared for output, the key becomes committed because the `snapshot_checker_(seq, kMaxSequence)` becomes true in the implementation of `KeyCommitted()`. Then `CompactionIterator` will try to zero its sequence number and hit assertion error if the key is a tombstone. To fix this issue, we should make the `CompactionIterator` see a consistent view of the input keys. Note that for write-prepared/write-unprepared, the background flush/compaction jobs already take a "job snapshot" before starting processing keys. The job snapshot is released only after the entire flush/compaction finishes. We can use this snapshot to determine whether a key is committed or not with minor change to `KeyCommitted()`. ``` inline bool KeyCommitted(SequenceNumber sequence) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(sequence, job_snapshot_) == SnapshotCheckerResult::kInSnapshot; } ``` As a result, whether a key is committed or not will remain a constant throughout compaction, causing no trouble for `CompactionIterator`s assertions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9830 Test Plan: make check Reviewed By: ltamasi Differential Revision: D35561162 Pulled By: riversand963 fbshipit-source-id: 0e00d200c195240341cfe6d34cbc86798b315b9f
3 years ago
const SnapshotChecker* snapshot_checker, JobContext* job_context,
std::shared_ptr<Cache> table_cache, EventLogger* event_logger,
bool paranoid_file_checks, bool measure_io_stats, const std::string& dbname,
CompactionJobStats* compaction_job_stats, Env::Priority thread_pri,
const std::shared_ptr<IOTracer>& io_tracer,
const std::atomic<bool>& manual_compaction_canceled,
const std::string& db_id, const std::string& db_session_id,
std::string full_history_ts_low, std::string trim_ts,
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
BlobFileCompletionCallback* blob_callback, int* bg_compaction_scheduled,
int* bg_bottom_compaction_scheduled)
: compact_(new CompactionState(compaction)),
compaction_stats_(compaction->compaction_reason(), 1),
db_options_(db_options),
mutable_db_options_copy_(mutable_db_options),
log_buffer_(log_buffer),
output_directory_(output_directory),
stats_(stats),
bottommost_level_(false),
write_hint_(Env::WLTH_NOT_SET),
compaction_job_stats_(compaction_job_stats),
job_id_(job_id),
dbname_(dbname),
db_id_(db_id),
db_session_id_(db_session_id),
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
file_options_(file_options),
env_(db_options.env),
io_tracer_(io_tracer),
fs_(db_options.fs, io_tracer),
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
file_options_for_read_(
fs_->OptimizeForCompactionTableRead(file_options, db_options_)),
versions_(versions),
shutting_down_(shutting_down),
manual_compaction_canceled_(manual_compaction_canceled),
db_directory_(db_directory),
blob_output_directory_(blob_output_directory),
db_mutex_(db_mutex),
db_error_handler_(db_error_handler),
existing_snapshots_(std::move(existing_snapshots)),
earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
snapshot_checker_(snapshot_checker),
CompactionIterator sees consistent view of which keys are committed (#9830) Summary: **This PR does not affect the functionality of `DB` and write-committed transactions.** `CompactionIterator` uses `KeyCommitted(seq)` to determine if a key in the database is committed. As the name 'write-committed' implies, if write-committed policy is used, a key exists in the database only if it is committed. In fact, the implementation of `KeyCommitted()` is as follows: ``` inline bool KeyCommitted(SequenceNumber seq) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(seq, kMaxSequence) == SnapshotCheckerResult::kInSnapshot; } ``` With that being said, we focus on write-prepared/write-unprepared transactions. A few notes: - A key can exist in the db even if it's uncommitted. Therefore, we rely on `snapshot_checker_` to determine data visibility. We also require that all writes go through transaction API instead of the raw `WriteBatch` + `Write`, thus at most one uncommitted version of one user key can exist in the database. - `CompactionIterator` outputs a key as long as the key is uncommitted. Due to the above reasons, it is possible that `CompactionIterator` decides to output an uncommitted key without doing further checks on the key (`NextFromInput()`). By the time the key is being prepared for output, the key becomes committed because the `snapshot_checker_(seq, kMaxSequence)` becomes true in the implementation of `KeyCommitted()`. Then `CompactionIterator` will try to zero its sequence number and hit assertion error if the key is a tombstone. To fix this issue, we should make the `CompactionIterator` see a consistent view of the input keys. Note that for write-prepared/write-unprepared, the background flush/compaction jobs already take a "job snapshot" before starting processing keys. The job snapshot is released only after the entire flush/compaction finishes. We can use this snapshot to determine whether a key is committed or not with minor change to `KeyCommitted()`. ``` inline bool KeyCommitted(SequenceNumber sequence) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(sequence, job_snapshot_) == SnapshotCheckerResult::kInSnapshot; } ``` As a result, whether a key is committed or not will remain a constant throughout compaction, causing no trouble for `CompactionIterator`s assertions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9830 Test Plan: make check Reviewed By: ltamasi Differential Revision: D35561162 Pulled By: riversand963 fbshipit-source-id: 0e00d200c195240341cfe6d34cbc86798b315b9f
3 years ago
job_context_(job_context),
table_cache_(std::move(table_cache)),
event_logger_(event_logger),
paranoid_file_checks_(paranoid_file_checks),
measure_io_stats_(measure_io_stats),
thread_pri_(thread_pri),
full_history_ts_low_(std::move(full_history_ts_low)),
trim_ts_(std::move(trim_ts)),
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
blob_callback_(blob_callback),
extra_num_subcompaction_threads_reserved_(0),
bg_compaction_scheduled_(bg_compaction_scheduled),
bg_bottom_compaction_scheduled_(bg_bottom_compaction_scheduled) {
assert(compaction_job_stats_ != nullptr);
assert(log_buffer_ != nullptr);
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
const auto* cfd = compact_->compaction->column_family_data();
ThreadStatusUtil::SetColumnFamily(cfd, cfd->ioptions()->env,
db_options_.enable_thread_tracking);
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
ThreadStatusUtil::SetThreadOperation(ThreadStatus::OP_COMPACTION);
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ReportStartedCompaction(compaction);
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
}
CompactionJob::~CompactionJob() {
assert(compact_ == nullptr);
ThreadStatusUtil::ResetThreadStatus();
}
void CompactionJob::ReportStartedCompaction(Compaction* compaction) {
const auto* cfd = compact_->compaction->column_family_data();
ThreadStatusUtil::SetColumnFamily(cfd, cfd->ioptions()->env,
db_options_.enable_thread_tracking);
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ThreadStatusUtil::SetThreadOperationProperty(ThreadStatus::COMPACTION_JOB_ID,
job_id_);
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ThreadStatusUtil::SetThreadOperationProperty(
ThreadStatus::COMPACTION_INPUT_OUTPUT_LEVEL,
(static_cast<uint64_t>(compact_->compaction->start_level()) << 32) +
compact_->compaction->output_level());
// In the current design, a CompactionJob is always created
// for non-trivial compaction.
assert(compaction->IsTrivialMove() == false ||
compaction->is_manual_compaction() == true);
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ThreadStatusUtil::SetThreadOperationProperty(
ThreadStatus::COMPACTION_PROP_FLAGS,
compaction->is_manual_compaction() +
(compaction->deletion_compaction() << 1));
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ThreadStatusUtil::SetThreadOperationProperty(
ThreadStatus::COMPACTION_TOTAL_INPUT_BYTES,
compaction->CalculateTotalInputSize());
IOSTATS_RESET(bytes_written);
IOSTATS_RESET(bytes_read);
ThreadStatusUtil::SetThreadOperationProperty(
ThreadStatus::COMPACTION_BYTES_WRITTEN, 0);
ThreadStatusUtil::SetThreadOperationProperty(
ThreadStatus::COMPACTION_BYTES_READ, 0);
// Set the thread operation after operation properties
// to ensure GetThreadList() can always show them all together.
ThreadStatusUtil::SetThreadOperation(ThreadStatus::OP_COMPACTION);
compaction_job_stats_->is_manual_compaction =
compaction->is_manual_compaction();
compaction_job_stats_->is_full_compaction = compaction->is_full_compaction();
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
}
void CompactionJob::Prepare() {
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_PREPARE);
// Generate file_levels_ for compaction before making Iterator
auto* c = compact_->compaction;
ColumnFamilyData* cfd = c->column_family_data();
assert(cfd != nullptr);
assert(cfd->current()->storage_info()->NumLevelFiles(
compact_->compaction->level()) > 0);
write_hint_ = cfd->CalculateSSTWriteHint(c->output_level());
bottommost_level_ = c->bottommost_level();
if (c->ShouldFormSubcompactions()) {
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
StopWatch sw(db_options_.clock, stats_, SUBCOMPACTION_SETUP_TIME);
GenSubcompactionBoundaries();
}
if (boundaries_.size() > 1) {
for (size_t i = 0; i <= boundaries_.size(); i++) {
compact_->sub_compact_states.emplace_back(
c, (i != 0) ? std::optional<Slice>(boundaries_[i - 1]) : std::nullopt,
(i != boundaries_.size()) ? std::optional<Slice>(boundaries_[i])
: std::nullopt,
static_cast<uint32_t>(i));
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
// assert to validate that boundaries don't have same user keys (without
// timestamp part).
assert(i == 0 || i == boundaries_.size() ||
cfd->user_comparator()->CompareWithoutTimestamp(
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
boundaries_[i - 1], boundaries_[i]) < 0);
}
RecordInHistogram(stats_, NUM_SUBCOMPACTIONS_SCHEDULED,
compact_->sub_compact_states.size());
} else {
compact_->sub_compact_states.emplace_back(c, std::nullopt, std::nullopt,
/*sub_job_id*/ 0);
}
// collect all seqno->time information from the input files which will be used
// to encode seqno->time to the output files.
uint64_t preserve_time_duration =
std::max(c->immutable_options()->preserve_internal_time_seconds,
c->immutable_options()->preclude_last_level_data_seconds);
if (preserve_time_duration > 0) {
// setup seqno_time_mapping_
seqno_time_mapping_.SetMaxTimeDuration(preserve_time_duration);
for (const auto& each_level : *c->inputs()) {
for (const auto& fmd : each_level.files) {
std::shared_ptr<const TableProperties> tp;
Status s = cfd->current()->GetTableProperties(&tp, fmd, nullptr);
if (s.ok()) {
seqno_time_mapping_.Add(tp->seqno_to_time_mapping)
.PermitUncheckedError();
seqno_time_mapping_.Add(fmd->fd.smallest_seqno,
fmd->oldest_ancester_time);
}
}
}
auto status = seqno_time_mapping_.Sort();
if (!status.ok()) {
ROCKS_LOG_WARN(db_options_.info_log,
"Invalid sequence number to time mapping: Status: %s",
status.ToString().c_str());
}
int64_t _current_time = 0;
status = db_options_.clock->GetCurrentTime(&_current_time);
if (!status.ok()) {
ROCKS_LOG_WARN(db_options_.info_log,
"Failed to get current time in compaction: Status: %s",
status.ToString().c_str());
// preserve all time information
preserve_time_min_seqno_ = 0;
preclude_last_level_min_seqno_ = 0;
} else {
seqno_time_mapping_.TruncateOldEntries(_current_time);
uint64_t preserve_time =
static_cast<uint64_t>(_current_time) > preserve_time_duration
? _current_time - preserve_time_duration
: 0;
preserve_time_min_seqno_ =
seqno_time_mapping_.GetOldestSequenceNum(preserve_time);
if (c->immutable_options()->preclude_last_level_data_seconds > 0) {
uint64_t preclude_last_level_time =
static_cast<uint64_t>(_current_time) >
c->immutable_options()->preclude_last_level_data_seconds
? _current_time -
c->immutable_options()->preclude_last_level_data_seconds
: 0;
preclude_last_level_min_seqno_ =
seqno_time_mapping_.GetOldestSequenceNum(preclude_last_level_time);
}
}
}
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
}
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
uint64_t CompactionJob::GetSubcompactionsLimit() {
return extra_num_subcompaction_threads_reserved_ +
std::max(
std::uint64_t(1),
static_cast<uint64_t>(compact_->compaction->max_subcompactions()));
}
void CompactionJob::AcquireSubcompactionResources(
int num_extra_required_subcompactions) {
TEST_SYNC_POINT("CompactionJob::AcquireSubcompactionResources:0");
TEST_SYNC_POINT("CompactionJob::AcquireSubcompactionResources:1");
int max_db_compactions =
DBImpl::GetBGJobLimits(
mutable_db_options_copy_.max_background_flushes,
mutable_db_options_copy_.max_background_compactions,
mutable_db_options_copy_.max_background_jobs,
versions_->GetColumnFamilySet()
->write_controller()
->NeedSpeedupCompaction())
.max_compactions;
Add missing mutex when reading from shared variable bg_bottom_compaction_scheduled_, bg_compaction_scheduled_ (#10610) Summary: **Context/Summary:** According to https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_job.h#L328-L332, any reading in the form of `*bg_compaction_scheduled_` , `*bg_bottom_compaction_scheduled_` should be protected by mutex, which isn't the case for some assert statement. This leads to a data race that can be repro-ed by the following command (command coming soon) ``` db=/dev/shm/rocksdb_crashtest_blackbox exp=/dev/shm/rocksdb_crashtest_expected rm -rf $db $exp mkdir -p $exp ./db_stress --clear_column_family_one_in=0 --column_families=1 --db=$db --delpercent=10 --delrangepercent=0 --destroy_db_initially=1 --expected_values_dir=$exp --iterpercent=0 --key_len_percent_dist=1,30,69 --max_key=1000000 --max_key_len=3 --prefixpercent=0 --readpercent=0 --reopen=0 --ops_per_thread=100000000 --value_size_mult=32 --writepercent=90 --compaction_pri=4 --use_txn=1 --level_compaction_dynamic_level_bytes=True --compaction_ttl=0 --compact_files_one_in=1000000 --compact_range_one_in=1000000 --value_size_mult=32 --verify_db_one_in=1000 --write_buffer_size=65536 --mark_for_compaction_one_file_in=10 --max_background_compactions=20 --max_key=25000000 --max_key_len=3 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=2097152 --target_file_size_base=2097152 --target_file_size_multiplier=2 ``` ``` WARNING: ThreadSanitizer: data race (pid=73424) Read of size 4 at 0x7b8c0000151c by thread T13: #0 ReleaseSubcompactionResources internal_repo_rocksdb/repo/db/compaction/compaction_job.cc:390 (db_stress+0x630aa3) https://github.com/facebook/rocksdb/issues/1 rocksdb::CompactionJob::Run() internal_repo_rocksdb/repo/db/compaction/compaction_job.cc:741 (db_stress+0x630aa3) https://github.com/facebook/rocksdb/issues/2 rocksdb::DBImpl::BackgroundCompaction(bool*, rocksdb::JobContext*, rocksdb::LogBuffer*, rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:3436 (db_stress+0x60b2cc) https://github.com/facebook/rocksdb/issues/3 rocksdb::DBImpl::BackgroundCallCompaction(rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2950 (db_stress+0x606d79) https://github.com/facebook/rocksdb/issues/4 rocksdb::DBImpl::BGWorkCompaction(void*) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2693 (db_stress+0x60356a) Previous write of size 4 at 0x7b8c0000151c by thread T12 (mutexes: write M438955329917552448): #0 rocksdb::DBImpl::BackgroundCallCompaction(rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:3018 (db_stress+0x6072a1) https://github.com/facebook/rocksdb/issues/1 rocksdb::DBImpl::BGWorkCompaction(void*) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2693 (db_stress+0x60356a) Location is heap block of size 6720 at 0x7b8c00000000 allocated by main thread: #0 operator new(unsigned long, std::align_val_t) <null> (db_stress+0xbab5bb) https://github.com/facebook/rocksdb/issues/1 rocksdb::DBImpl::Open(rocksdb::DBOptions const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::vector<rocksdb::ColumnFamilyDescriptor, std::allocator<rocksdb::ColumnFamilyDescriptor> > const&, std::vector<rocksdb::ColumnFamilyHandle*, std::allocator<rocksdb::ColumnFamilyHandle*> >*, rocksdb::DB**, bool, bool) internal_repo_rocksdb/repo/db/db_impl/db_impl_open.cc:1811 (db_stress+0x69769a) https://github.com/facebook/rocksdb/issues/2 rocksdb::TransactionDB::Open(rocksdb::DBOptions const&, rocksdb::TransactionDBOptions const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::vector<rocksdb::ColumnFamilyDescriptor, std::allocator<rocksdb::ColumnFamilyDescriptor> > const&, std::vector<rocksdb::ColumnFamilyHandle*, std::allocator<rocksdb::ColumnFamilyHandle*> >*, rocksdb::TransactionDB**) internal_repo_rocksdb/repo/utilities/transactions/pessimistic_transaction_db.cc:258 (db_stress+0x8ae1f4) https://github.com/facebook/rocksdb/issues/3 rocksdb::StressTest::Open(rocksdb::SharedState*) internal_repo_rocksdb/repo/db_stress_tool/db_stress_test_base.cc:2611 (db_stress+0x32b927) https://github.com/facebook/rocksdb/issues/4 rocksdb::StressTest::InitDb(rocksdb::SharedState*) internal_repo_rocksdb/repo/db_stress_tool/db_stress_test_base.cc:290 (db_stress+0x34712c) ``` This PR added all the missing mutex that should've been in place Pull Request resolved: https://github.com/facebook/rocksdb/pull/10610 Test Plan: - Past repro command - Existing CI Reviewed By: riversand963 Differential Revision: D39143016 Pulled By: hx235 fbshipit-source-id: 51dd4db55ad306f3dbda5d0dd54d6f2513cf70f2
2 years ago
InstrumentedMutexLock l(db_mutex_);
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
// Apply min function first since We need to compute the extra subcompaction
// against compaction limits. And then try to reserve threads for extra
// subcompactions. The actual number of reserved threads could be less than
// the desired number.
int available_bg_compactions_against_db_limit =
std::max(max_db_compactions - *bg_compaction_scheduled_ -
*bg_bottom_compaction_scheduled_,
0);
// Reservation only supports backgrdoun threads of which the priority is
// between BOTTOM and HIGH. Need to degrade the priority to HIGH if the
// origin thread_pri_ is higher than that. Similar to ReleaseThreads().
extra_num_subcompaction_threads_reserved_ =
env_->ReserveThreads(std::min(num_extra_required_subcompactions,
available_bg_compactions_against_db_limit),
std::min(thread_pri_, Env::Priority::HIGH));
// Update bg_compaction_scheduled_ or bg_bottom_compaction_scheduled_
// depending on if this compaction has the bottommost priority
if (thread_pri_ == Env::Priority::BOTTOM) {
*bg_bottom_compaction_scheduled_ +=
extra_num_subcompaction_threads_reserved_;
} else {
*bg_compaction_scheduled_ += extra_num_subcompaction_threads_reserved_;
}
}
void CompactionJob::ShrinkSubcompactionResources(uint64_t num_extra_resources) {
// Do nothing when we have zero resources to shrink
if (num_extra_resources == 0) return;
db_mutex_->Lock();
// We cannot release threads more than what we reserved before
int extra_num_subcompaction_threads_released = env_->ReleaseThreads(
(int)num_extra_resources, std::min(thread_pri_, Env::Priority::HIGH));
// Update the number of reserved threads and the number of background
// scheduled compactions for this compaction job
extra_num_subcompaction_threads_reserved_ -=
extra_num_subcompaction_threads_released;
// TODO (zichen): design a test case with new subcompaction partitioning
// when the number of actual partitions is less than the number of planned
// partitions
assert(extra_num_subcompaction_threads_released == (int)num_extra_resources);
// Update bg_compaction_scheduled_ or bg_bottom_compaction_scheduled_
// depending on if this compaction has the bottommost priority
if (thread_pri_ == Env::Priority::BOTTOM) {
*bg_bottom_compaction_scheduled_ -=
extra_num_subcompaction_threads_released;
} else {
*bg_compaction_scheduled_ -= extra_num_subcompaction_threads_released;
}
db_mutex_->Unlock();
TEST_SYNC_POINT("CompactionJob::ShrinkSubcompactionResources:0");
}
void CompactionJob::ReleaseSubcompactionResources() {
if (extra_num_subcompaction_threads_reserved_ == 0) {
return;
}
Add missing mutex when reading from shared variable bg_bottom_compaction_scheduled_, bg_compaction_scheduled_ (#10610) Summary: **Context/Summary:** According to https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_job.h#L328-L332, any reading in the form of `*bg_compaction_scheduled_` , `*bg_bottom_compaction_scheduled_` should be protected by mutex, which isn't the case for some assert statement. This leads to a data race that can be repro-ed by the following command (command coming soon) ``` db=/dev/shm/rocksdb_crashtest_blackbox exp=/dev/shm/rocksdb_crashtest_expected rm -rf $db $exp mkdir -p $exp ./db_stress --clear_column_family_one_in=0 --column_families=1 --db=$db --delpercent=10 --delrangepercent=0 --destroy_db_initially=1 --expected_values_dir=$exp --iterpercent=0 --key_len_percent_dist=1,30,69 --max_key=1000000 --max_key_len=3 --prefixpercent=0 --readpercent=0 --reopen=0 --ops_per_thread=100000000 --value_size_mult=32 --writepercent=90 --compaction_pri=4 --use_txn=1 --level_compaction_dynamic_level_bytes=True --compaction_ttl=0 --compact_files_one_in=1000000 --compact_range_one_in=1000000 --value_size_mult=32 --verify_db_one_in=1000 --write_buffer_size=65536 --mark_for_compaction_one_file_in=10 --max_background_compactions=20 --max_key=25000000 --max_key_len=3 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=2097152 --target_file_size_base=2097152 --target_file_size_multiplier=2 ``` ``` WARNING: ThreadSanitizer: data race (pid=73424) Read of size 4 at 0x7b8c0000151c by thread T13: #0 ReleaseSubcompactionResources internal_repo_rocksdb/repo/db/compaction/compaction_job.cc:390 (db_stress+0x630aa3) https://github.com/facebook/rocksdb/issues/1 rocksdb::CompactionJob::Run() internal_repo_rocksdb/repo/db/compaction/compaction_job.cc:741 (db_stress+0x630aa3) https://github.com/facebook/rocksdb/issues/2 rocksdb::DBImpl::BackgroundCompaction(bool*, rocksdb::JobContext*, rocksdb::LogBuffer*, rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:3436 (db_stress+0x60b2cc) https://github.com/facebook/rocksdb/issues/3 rocksdb::DBImpl::BackgroundCallCompaction(rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2950 (db_stress+0x606d79) https://github.com/facebook/rocksdb/issues/4 rocksdb::DBImpl::BGWorkCompaction(void*) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2693 (db_stress+0x60356a) Previous write of size 4 at 0x7b8c0000151c by thread T12 (mutexes: write M438955329917552448): #0 rocksdb::DBImpl::BackgroundCallCompaction(rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:3018 (db_stress+0x6072a1) https://github.com/facebook/rocksdb/issues/1 rocksdb::DBImpl::BGWorkCompaction(void*) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2693 (db_stress+0x60356a) Location is heap block of size 6720 at 0x7b8c00000000 allocated by main thread: #0 operator new(unsigned long, std::align_val_t) <null> (db_stress+0xbab5bb) https://github.com/facebook/rocksdb/issues/1 rocksdb::DBImpl::Open(rocksdb::DBOptions const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::vector<rocksdb::ColumnFamilyDescriptor, std::allocator<rocksdb::ColumnFamilyDescriptor> > const&, std::vector<rocksdb::ColumnFamilyHandle*, std::allocator<rocksdb::ColumnFamilyHandle*> >*, rocksdb::DB**, bool, bool) internal_repo_rocksdb/repo/db/db_impl/db_impl_open.cc:1811 (db_stress+0x69769a) https://github.com/facebook/rocksdb/issues/2 rocksdb::TransactionDB::Open(rocksdb::DBOptions const&, rocksdb::TransactionDBOptions const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::vector<rocksdb::ColumnFamilyDescriptor, std::allocator<rocksdb::ColumnFamilyDescriptor> > const&, std::vector<rocksdb::ColumnFamilyHandle*, std::allocator<rocksdb::ColumnFamilyHandle*> >*, rocksdb::TransactionDB**) internal_repo_rocksdb/repo/utilities/transactions/pessimistic_transaction_db.cc:258 (db_stress+0x8ae1f4) https://github.com/facebook/rocksdb/issues/3 rocksdb::StressTest::Open(rocksdb::SharedState*) internal_repo_rocksdb/repo/db_stress_tool/db_stress_test_base.cc:2611 (db_stress+0x32b927) https://github.com/facebook/rocksdb/issues/4 rocksdb::StressTest::InitDb(rocksdb::SharedState*) internal_repo_rocksdb/repo/db_stress_tool/db_stress_test_base.cc:290 (db_stress+0x34712c) ``` This PR added all the missing mutex that should've been in place Pull Request resolved: https://github.com/facebook/rocksdb/pull/10610 Test Plan: - Past repro command - Existing CI Reviewed By: riversand963 Differential Revision: D39143016 Pulled By: hx235 fbshipit-source-id: 51dd4db55ad306f3dbda5d0dd54d6f2513cf70f2
2 years ago
{
InstrumentedMutexLock l(db_mutex_);
// The number of reserved threads becomes larger than 0 only if the
// compaction prioity is round robin and there is no sufficient
// sub-compactions available
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
Add missing mutex when reading from shared variable bg_bottom_compaction_scheduled_, bg_compaction_scheduled_ (#10610) Summary: **Context/Summary:** According to https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_job.h#L328-L332, any reading in the form of `*bg_compaction_scheduled_` , `*bg_bottom_compaction_scheduled_` should be protected by mutex, which isn't the case for some assert statement. This leads to a data race that can be repro-ed by the following command (command coming soon) ``` db=/dev/shm/rocksdb_crashtest_blackbox exp=/dev/shm/rocksdb_crashtest_expected rm -rf $db $exp mkdir -p $exp ./db_stress --clear_column_family_one_in=0 --column_families=1 --db=$db --delpercent=10 --delrangepercent=0 --destroy_db_initially=1 --expected_values_dir=$exp --iterpercent=0 --key_len_percent_dist=1,30,69 --max_key=1000000 --max_key_len=3 --prefixpercent=0 --readpercent=0 --reopen=0 --ops_per_thread=100000000 --value_size_mult=32 --writepercent=90 --compaction_pri=4 --use_txn=1 --level_compaction_dynamic_level_bytes=True --compaction_ttl=0 --compact_files_one_in=1000000 --compact_range_one_in=1000000 --value_size_mult=32 --verify_db_one_in=1000 --write_buffer_size=65536 --mark_for_compaction_one_file_in=10 --max_background_compactions=20 --max_key=25000000 --max_key_len=3 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=2097152 --target_file_size_base=2097152 --target_file_size_multiplier=2 ``` ``` WARNING: ThreadSanitizer: data race (pid=73424) Read of size 4 at 0x7b8c0000151c by thread T13: #0 ReleaseSubcompactionResources internal_repo_rocksdb/repo/db/compaction/compaction_job.cc:390 (db_stress+0x630aa3) https://github.com/facebook/rocksdb/issues/1 rocksdb::CompactionJob::Run() internal_repo_rocksdb/repo/db/compaction/compaction_job.cc:741 (db_stress+0x630aa3) https://github.com/facebook/rocksdb/issues/2 rocksdb::DBImpl::BackgroundCompaction(bool*, rocksdb::JobContext*, rocksdb::LogBuffer*, rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:3436 (db_stress+0x60b2cc) https://github.com/facebook/rocksdb/issues/3 rocksdb::DBImpl::BackgroundCallCompaction(rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2950 (db_stress+0x606d79) https://github.com/facebook/rocksdb/issues/4 rocksdb::DBImpl::BGWorkCompaction(void*) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2693 (db_stress+0x60356a) Previous write of size 4 at 0x7b8c0000151c by thread T12 (mutexes: write M438955329917552448): #0 rocksdb::DBImpl::BackgroundCallCompaction(rocksdb::DBImpl::PrepickedCompaction*, rocksdb::Env::Priority) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:3018 (db_stress+0x6072a1) https://github.com/facebook/rocksdb/issues/1 rocksdb::DBImpl::BGWorkCompaction(void*) internal_repo_rocksdb/repo/db/db_impl/db_impl_compaction_flush.cc:2693 (db_stress+0x60356a) Location is heap block of size 6720 at 0x7b8c00000000 allocated by main thread: #0 operator new(unsigned long, std::align_val_t) <null> (db_stress+0xbab5bb) https://github.com/facebook/rocksdb/issues/1 rocksdb::DBImpl::Open(rocksdb::DBOptions const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::vector<rocksdb::ColumnFamilyDescriptor, std::allocator<rocksdb::ColumnFamilyDescriptor> > const&, std::vector<rocksdb::ColumnFamilyHandle*, std::allocator<rocksdb::ColumnFamilyHandle*> >*, rocksdb::DB**, bool, bool) internal_repo_rocksdb/repo/db/db_impl/db_impl_open.cc:1811 (db_stress+0x69769a) https://github.com/facebook/rocksdb/issues/2 rocksdb::TransactionDB::Open(rocksdb::DBOptions const&, rocksdb::TransactionDBOptions const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::vector<rocksdb::ColumnFamilyDescriptor, std::allocator<rocksdb::ColumnFamilyDescriptor> > const&, std::vector<rocksdb::ColumnFamilyHandle*, std::allocator<rocksdb::ColumnFamilyHandle*> >*, rocksdb::TransactionDB**) internal_repo_rocksdb/repo/utilities/transactions/pessimistic_transaction_db.cc:258 (db_stress+0x8ae1f4) https://github.com/facebook/rocksdb/issues/3 rocksdb::StressTest::Open(rocksdb::SharedState*) internal_repo_rocksdb/repo/db_stress_tool/db_stress_test_base.cc:2611 (db_stress+0x32b927) https://github.com/facebook/rocksdb/issues/4 rocksdb::StressTest::InitDb(rocksdb::SharedState*) internal_repo_rocksdb/repo/db_stress_tool/db_stress_test_base.cc:290 (db_stress+0x34712c) ``` This PR added all the missing mutex that should've been in place Pull Request resolved: https://github.com/facebook/rocksdb/pull/10610 Test Plan: - Past repro command - Existing CI Reviewed By: riversand963 Differential Revision: D39143016 Pulled By: hx235 fbshipit-source-id: 51dd4db55ad306f3dbda5d0dd54d6f2513cf70f2
2 years ago
// The scheduled compaction must be no less than 1 + extra number
// subcompactions using acquired resources since this compaction job has not
// finished yet
assert(*bg_bottom_compaction_scheduled_ >=
1 + extra_num_subcompaction_threads_reserved_ ||
*bg_compaction_scheduled_ >=
1 + extra_num_subcompaction_threads_reserved_);
}
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
ShrinkSubcompactionResources(extra_num_subcompaction_threads_reserved_);
}
struct RangeWithSize {
Range range;
uint64_t size;
RangeWithSize(const Slice& a, const Slice& b, uint64_t s = 0)
: range(a, b), size(s) {}
};
void CompactionJob::GenSubcompactionBoundaries() {
// The goal is to find some boundary keys so that we can evenly partition
// the compaction input data into max_subcompactions ranges.
// For every input file, we ask TableReader to estimate 128 anchor points
// that evenly partition the input file into 128 ranges and the range
// sizes. This can be calculated by scanning index blocks of the file.
// Once we have the anchor points for all the input files, we merge them
// together and try to find keys dividing ranges evenly.
// For example, if we have two input files, and each returns following
// ranges:
// File1: (a1, 1000), (b1, 1200), (c1, 1100)
// File2: (a2, 1100), (b2, 1000), (c2, 1000)
// We total sort the keys to following:
// (a1, 1000), (a2, 1100), (b1, 1200), (b2, 1000), (c1, 1100), (c2, 1000)
// We calculate the total size by adding up all ranges' size, which is 6400.
// If we would like to partition into 2 subcompactions, the target of the
// range size is 3200. Based on the size, we take "b1" as the partition key
// since the first three ranges would hit 3200.
//
// Note that the ranges are actually overlapping. For example, in the example
// above, the range ending with "b1" is overlapping with the range ending with
// "b2". So the size 1000+1100+1200 is an underestimation of data size up to
// "b1". In extreme cases where we only compact N L0 files, a range can
// overlap with N-1 other ranges. Since we requested a relatively large number
// (128) of ranges from each input files, even N range overlapping would
// cause relatively small inaccuracy.
auto* c = compact_->compaction;
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
if (c->max_subcompactions() <= 1 &&
!(c->immutable_options()->compaction_pri == kRoundRobin &&
c->immutable_options()->compaction_style == kCompactionStyleLevel)) {
return;
}
auto* cfd = c->column_family_data();
const Comparator* cfd_comparator = cfd->user_comparator();
const InternalKeyComparator& icomp = cfd->internal_comparator();
auto* v = compact_->compaction->input_version();
int base_level = v->storage_info()->base_level();
InstrumentedMutexUnlock unlock_guard(db_mutex_);
uint64_t total_size = 0;
std::vector<TableReader::Anchor> all_anchors;
int start_lvl = c->start_level();
int out_lvl = c->output_level();
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
for (size_t lvl_idx = 0; lvl_idx < c->num_input_levels(); lvl_idx++) {
int lvl = c->level(lvl_idx);
if (lvl >= start_lvl && lvl <= out_lvl) {
const LevelFilesBrief* flevel = c->input_levels(lvl_idx);
size_t num_files = flevel->num_files;
if (num_files == 0) {
continue;
}
for (size_t i = 0; i < num_files; i++) {
FileMetaData* f = flevel->files[i].file_metadata;
std::vector<TableReader::Anchor> my_anchors;
Status s = cfd->table_cache()->ApproximateKeyAnchors(
Always verify SST unique IDs on SST file open (#10532) Summary: Although we've been tracking SST unique IDs in the DB manifest unconditionally, checking has been opt-in and with an extra pass at DB::Open time. This changes the behavior of `verify_sst_unique_id_in_manifest` to check unique ID against manifest every time an SST file is opened through table cache (normal DB operations), replacing the explicit pass over files at DB::Open time. This change also enables the option by default and removes the "EXPERIMENTAL" designation. One possible criticism is that the option no longer ensures the integrity of a DB at Open time. This is far from an all-or-nothing issue. Verifying the IDs of all SST files hardly ensures all the data in the DB is readable. (VerifyChecksum is supposed to do that.) Also, with max_open_files=-1 (default, extremely common), all SST files are opened at DB::Open time anyway. Implementation details: * `VerifySstUniqueIdInManifest()` functions are the extra/explicit pass that is now removed. * Unit tests that manipulate/corrupt table properties have to opt out of this check, because that corrupts the "actual" unique id. (And even for testing we don't currently have a mechanism to set "no unique id" in the in-memory file metadata for new files.) * A lot of other unit test churn relates to (a) default checking on, and (b) checking on SST open even without DB::Open (e.g. on flush) * Use `FileMetaData` for more `TableCache` operations (in place of `FileDescriptor`) so that we have access to the unique_id whenever we might need to open an SST file. **There is the possibility of performance impact because we can no longer use the more localized `fd` part of an `FdWithKeyRange` but instead follow the `file_metadata` pointer. However, this change (possible regression) is only done for `GetMemoryUsageByTableReaders`.** * Removed a completely unnecessary constructor overload of `TableReaderOptions` Possible follow-up: * Verification only happens when opening through table cache. Are there more places where this should happen? * Improve error message when there is a file size mismatch vs. manifest (FIXME added in the appropriate place). * I'm not sure there's a justification for `FileDescriptor` to be distinct from `FileMetaData`. * I'm skeptical that `FdWithKeyRange` really still makes sense for optimizing some data locality by duplicating some data in memory, but I could be wrong. * An unnecessary overload of NewTableReader was recently added, in the public API nonetheless (though unusable there). It should be cleaned up to put most things under `TableReaderOptions`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10532 Test Plan: updated unit tests Performance test showing no significant difference (just noise I think): `./db_bench -benchmarks=readwhilewriting[-X10] -num=3000000 -disable_wal=1 -bloom_bits=8 -write_buffer_size=1000000 -target_file_size_base=1000000` Before: readwhilewriting [AVG 10 runs] : 68702 (± 6932) ops/sec After: readwhilewriting [AVG 10 runs] : 68239 (± 7198) ops/sec Reviewed By: jay-zhuang Differential Revision: D38765551 Pulled By: pdillinger fbshipit-source-id: a827a708155f12344ab2a5c16e7701c7636da4c2
2 years ago
ReadOptions(), icomp, *f, my_anchors);
if (!s.ok() || my_anchors.empty()) {
my_anchors.emplace_back(f->largest.user_key(), f->fd.GetFileSize());
}
for (auto& ac : my_anchors) {
// Can be optimize to avoid this loop.
total_size += ac.range_size;
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
}
all_anchors.insert(all_anchors.end(), my_anchors.begin(),
my_anchors.end());
}
}
}
// Here we total sort all the anchor points across all files and go through
// them in the sorted order to find partitioning boundaries.
// Not the most efficient implementation. A much more efficient algorithm
// probably exists. But they are more complex. If performance turns out to
// be a problem, we can optimize.
std::sort(
all_anchors.begin(), all_anchors.end(),
[cfd_comparator](TableReader::Anchor& a, TableReader::Anchor& b) -> bool {
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
return cfd_comparator->CompareWithoutTimestamp(a.user_key, b.user_key) <
0;
});
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
// Remove duplicated entries from boundaries.
all_anchors.erase(
std::unique(all_anchors.begin(), all_anchors.end(),
[cfd_comparator](TableReader::Anchor& a,
TableReader::Anchor& b) -> bool {
return cfd_comparator->CompareWithoutTimestamp(
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
a.user_key, b.user_key) == 0;
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
}),
all_anchors.end());
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
// Get the number of planned subcompactions, may update reserve threads
// and update extra_num_subcompaction_threads_reserved_ for round-robin
uint64_t num_planned_subcompactions;
if (c->immutable_options()->compaction_pri == kRoundRobin &&
c->immutable_options()->compaction_style == kCompactionStyleLevel) {
// For round-robin compaction prioity, we need to employ more
// subcompactions (may exceed the max_subcompaction limit). The extra
// subcompactions will be executed using reserved threads and taken into
// account bg_compaction_scheduled or bg_bottom_compaction_scheduled.
// Initialized by the number of input files
num_planned_subcompactions = static_cast<uint64_t>(c->num_input_files(0));
uint64_t max_subcompactions_limit = GetSubcompactionsLimit();
if (max_subcompactions_limit < num_planned_subcompactions) {
// Assert two pointers are not empty so that we can use extra
// subcompactions against db compaction limits
assert(bg_bottom_compaction_scheduled_ != nullptr);
assert(bg_compaction_scheduled_ != nullptr);
// Reserve resources when max_subcompaction is not sufficient
AcquireSubcompactionResources(
(int)(num_planned_subcompactions - max_subcompactions_limit));
// Subcompactions limit changes after acquiring additional resources.
// Need to call GetSubcompactionsLimit() again to update the number
// of planned subcompactions
num_planned_subcompactions =
std::min(num_planned_subcompactions, GetSubcompactionsLimit());
} else {
num_planned_subcompactions = max_subcompactions_limit;
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
}
} else {
num_planned_subcompactions = GetSubcompactionsLimit();
}
TEST_SYNC_POINT_CALLBACK("CompactionJob::GenSubcompactionBoundaries:0",
&num_planned_subcompactions);
if (num_planned_subcompactions == 1) return;
// Group the ranges into subcompactions
uint64_t target_range_size = std::max(
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
total_size / num_planned_subcompactions,
MaxFileSizeForLevel(
*(c->mutable_cf_options()), out_lvl,
c->immutable_options()->compaction_style, base_level,
c->immutable_options()->level_compaction_dynamic_level_bytes));
if (target_range_size >= total_size) {
return;
}
uint64_t next_threshold = target_range_size;
uint64_t cumulative_size = 0;
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
uint64_t num_actual_subcompactions = 1U;
for (TableReader::Anchor& anchor : all_anchors) {
cumulative_size += anchor.range_size;
if (cumulative_size > next_threshold) {
next_threshold += target_range_size;
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
num_actual_subcompactions++;
boundaries_.push_back(anchor.user_key);
}
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
if (num_actual_subcompactions == num_planned_subcompactions) {
break;
}
}
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
TEST_SYNC_POINT_CALLBACK("CompactionJob::GenSubcompactionBoundaries:1",
&num_actual_subcompactions);
// Shrink extra subcompactions resources when extra resrouces are acquired
ShrinkSubcompactionResources(
std::min((int)(num_planned_subcompactions - num_actual_subcompactions),
extra_num_subcompaction_threads_reserved_));
}
Status CompactionJob::Run() {
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_RUN);
TEST_SYNC_POINT("CompactionJob::Run():Start");
log_buffer_->FlushBufferToLog();
LogCompaction();
const size_t num_threads = compact_->sub_compact_states.size();
assert(num_threads > 0);
const uint64_t start_micros = db_options_.clock->NowMicros();
// Launch a thread for each of subcompactions 1...num_threads-1
std::vector<port::Thread> thread_pool;
thread_pool.reserve(num_threads - 1);
for (size_t i = 1; i < compact_->sub_compact_states.size(); i++) {
thread_pool.emplace_back(&CompactionJob::ProcessKeyValueCompaction, this,
&compact_->sub_compact_states[i]);
}
// Always schedule the first subcompaction (whether or not there are also
// others) in the current thread to be efficient with resources
ProcessKeyValueCompaction(&compact_->sub_compact_states[0]);
// Wait for all other threads (if there are any) to finish execution
for (auto& thread : thread_pool) {
thread.join();
}
compaction_stats_.SetMicros(db_options_.clock->NowMicros() - start_micros);
for (auto& state : compact_->sub_compact_states) {
compaction_stats_.AddCpuMicros(state.compaction_job_stats.cpu_micros);
state.RemoveLastEmptyOutput();
}
RecordTimeToHistogram(stats_, COMPACTION_TIME,
compaction_stats_.stats.micros);
RecordTimeToHistogram(stats_, COMPACTION_CPU_TIME,
compaction_stats_.stats.cpu_micros);
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
TEST_SYNC_POINT("CompactionJob::Run:BeforeVerify");
// Check if any thread encountered an error during execution
Status status;
IOStatus io_s;
bool wrote_new_blob_files = false;
for (const auto& state : compact_->sub_compact_states) {
if (!state.status.ok()) {
status = state.status;
io_s = state.io_status;
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
break;
}
if (state.Current().HasBlobFileAdditions()) {
wrote_new_blob_files = true;
}
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
}
if (io_status_.ok()) {
io_status_ = io_s;
}
if (status.ok()) {
constexpr IODebugContext* dbg = nullptr;
if (output_directory_) {
io_s = output_directory_->FsyncWithDirOptions(
IOOptions(), dbg,
DirFsyncOptions(DirFsyncOptions::FsyncReason::kNewFileSynced));
}
if (io_s.ok() && wrote_new_blob_files && blob_output_directory_ &&
blob_output_directory_ != output_directory_) {
io_s = blob_output_directory_->FsyncWithDirOptions(
IOOptions(), dbg,
DirFsyncOptions(DirFsyncOptions::FsyncReason::kNewFileSynced));
}
}
if (io_status_.ok()) {
io_status_ = io_s;
}
if (status.ok()) {
status = io_s;
}
if (status.ok()) {
thread_pool.clear();
std::vector<const CompactionOutputs::Output*> files_output;
for (const auto& state : compact_->sub_compact_states) {
for (const auto& output : state.GetOutputs()) {
files_output.emplace_back(&output);
}
}
ColumnFamilyData* cfd = compact_->compaction->column_family_data();
Fast path for detecting unchanged prefix_extractor (#9407) Summary: Fixes a major performance regression in 6.26, where extra CPU is spent in SliceTransform::AsString when reads involve a prefix_extractor (Get, MultiGet, Seek). Common case performance is now better than 6.25. This change creates a "fast path" for verifying that the current prefix extractor is unchanged and compatible with what was used to generate a table file. This fast path detects the common case by pointer comparison on the current prefix_extractor and a "known good" prefix extractor (if applicable) that is saved at the time the table reader is opened. The "known good" prefix extractor is saved as another shared_ptr copy (in an existing field, however) to ensure the pointer is not recycled. When the prefix_extractor has changed to a different instance but same compatible configuration (rare, odd), performance is still a regression compared to 6.25, but this is likely acceptable because of the oddity of such a case. The performance of incompatible prefix_extractor is essentially unchanged. Also fixed a minor case (ForwardIterator) where a prefix_extractor could be used via a raw pointer after being freed as a shared_ptr, if replaced via SetOptions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9407 Test Plan: ## Performance Populate DB with `TEST_TMPDIR=/dev/shm/rocksdb ./db_bench -benchmarks=fillrandom -num=10000000 -disable_wal=1 -write_buffer_size=10000000 -bloom_bits=16 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=0 -prefix_size=12` Running head-to-head comparisons simultaneously with `TEST_TMPDIR=/dev/shm/rocksdb ./db_bench -use_existing_db -readonly -benchmarks=seekrandom -num=10000000 -duration=20 -disable_wal=1 -bloom_bits=16 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=0 -prefix_size=12` Below each is compared by ops/sec vs. baseline which is version 6.25 (multiple baseline runs because of variable machine load) v6.26: 4833 vs. 6698 (<- major regression!) v6.27: 4737 vs. 6397 (still) New: 6704 vs. 6461 (better than baseline in common case) Disabled fastpath: 4843 vs. 6389 (e.g. if prefix extractor instance changes but is still compatible) Changed prefix size (no usable filter) in new: 787 vs. 5927 Changed prefix size (no usable filter) in new & baseline: 773 vs. 784 Reviewed By: mrambacher Differential Revision: D33677812 Pulled By: pdillinger fbshipit-source-id: 571d9711c461fb97f957378a061b7e7dbc4d6a76
3 years ago
auto& prefix_extractor =
compact_->compaction->mutable_cf_options()->prefix_extractor;
std::atomic<size_t> next_file_idx(0);
auto verify_table = [&](Status& output_status) {
while (true) {
size_t file_idx = next_file_idx.fetch_add(1);
if (file_idx >= files_output.size()) {
break;
}
// Verify that the table is usable
// We set for_compaction to false and don't
// OptimizeForCompactionTableRead here because this is a special case
// after we finish the table building No matter whether
// use_direct_io_for_flush_and_compaction is true, we will regard this
// verification as user reads since the goal is to cache it here for
// further user reads
ReadOptions read_options;
InternalIterator* iter = cfd->table_cache()->NewIterator(
read_options, file_options_, cfd->internal_comparator(),
files_output[file_idx]->meta, /*range_del_agg=*/nullptr,
prefix_extractor,
/*table_reader_ptr=*/nullptr,
cfd->internal_stats()->GetFileReadHist(
compact_->compaction->output_level()),
TableReaderCaller::kCompactionRefill, /*arena=*/nullptr,
/*skip_filters=*/false, compact_->compaction->output_level(),
MaxFileSizeForL0MetaPin(
*compact_->compaction->mutable_cf_options()),
/*smallest_compaction_key=*/nullptr,
Properly report IO errors when IndexType::kBinarySearchWithFirstKey is used (#6621) Summary: Context: Index type `kBinarySearchWithFirstKey` added the ability for sst file iterator to sometimes report a key from index without reading the corresponding data block. This is useful when sst blocks are cut at some meaningful boundaries (e.g. one block per key prefix), and many seeks land between blocks (e.g. for each prefix, the ranges of keys in different sst files are nearly disjoint, so a typical seek needs to read a data block from only one file even if all files have the prefix). But this added a new error condition, which rocksdb code was really not equipped to deal with: `InternalIterator::value()` may fail with an IO error or Status::Incomplete, but it's just a method returning a Slice, with no way to report error instead. Before this PR, this type of error wasn't handled at all (an empty slice was returned), and kBinarySearchWithFirstKey implementation was considered a prototype. Now that we (LogDevice) have experimented with kBinarySearchWithFirstKey for a while and confirmed that it's really useful, this PR is adding the missing error handling. It's a pretty inconvenient situation implementation-wise. The error needs to be reported from InternalIterator when trying to access value. But there are ~700 call sites of `InternalIterator::value()`, most of which either can't hit the error condition (because the iterator is reading from memtable or from index or something) or wouldn't benefit from the deferred loading of the value (e.g. compaction iterator that reads all values anyway). Adding error handling to all these call sites would needlessly bloat the code. So instead I made the deferred value loading optional: only the call sites that may use deferred loading have to call the new method `PrepareValue()` before calling `value()`. The feature is enabled with a new bool argument `allow_unprepared_value` to a bunch of methods that create iterators (it wouldn't make sense to put it in ReadOptions because it's completely internal to iterators, with virtually no user-visible effect). Lmk if you have better ideas. Note that the deferred value loading only happens for *internal* iterators. The user-visible iterator (DBIter) always prepares the value before returning from Seek/Next/etc. We could go further and add an API to defer that value loading too, but that's most likely not useful for LogDevice, so it doesn't seem worth the complexity for now. Pull Request resolved: https://github.com/facebook/rocksdb/pull/6621 Test Plan: make -j5 check . Will also deploy to some logdevice test clusters and look at stats. Reviewed By: siying Differential Revision: D20786930 Pulled By: al13n321 fbshipit-source-id: 6da77d918bad3780522e918f17f4d5513d3e99ee
5 years ago
/*largest_compaction_key=*/nullptr,
/*allow_unprepared_value=*/false);
auto s = iter->status();
if (s.ok() && paranoid_file_checks_) {
OutputValidator validator(cfd->internal_comparator(),
/*_enable_order_check=*/true,
/*_enable_hash=*/true);
for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
s = validator.Add(iter->key(), iter->value());
if (!s.ok()) {
break;
}
}
if (s.ok()) {
s = iter->status();
}
if (s.ok() &&
!validator.CompareValidator(files_output[file_idx]->validator)) {
s = Status::Corruption("Paranoid checksums do not match");
}
}
delete iter;
if (!s.ok()) {
output_status = s;
break;
}
}
};
for (size_t i = 1; i < compact_->sub_compact_states.size(); i++) {
thread_pool.emplace_back(
verify_table, std::ref(compact_->sub_compact_states[i].status));
}
verify_table(compact_->sub_compact_states[0].status);
for (auto& thread : thread_pool) {
thread.join();
}
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
for (const auto& state : compact_->sub_compact_states) {
if (!state.status.ok()) {
status = state.status;
break;
}
}
}
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
ReleaseSubcompactionResources();
TEST_SYNC_POINT("CompactionJob::ReleaseSubcompactionResources:0");
TEST_SYNC_POINT("CompactionJob::ReleaseSubcompactionResources:1");
TablePropertiesCollection tp;
for (const auto& state : compact_->sub_compact_states) {
for (const auto& output : state.GetOutputs()) {
auto fn =
TableFileName(state.compaction->immutable_options()->cf_paths,
output.meta.fd.GetNumber(), output.meta.fd.GetPathId());
tp[fn] = output.table_properties;
}
}
compact_->compaction->SetOutputTableProperties(std::move(tp));
// Finish up all book-keeping to unify the subcompaction results
compact_->AggregateCompactionStats(compaction_stats_, *compaction_job_stats_);
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
UpdateCompactionStats();
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
RecordCompactionIOStats();
LogFlush(db_options_.info_log);
TEST_SYNC_POINT("CompactionJob::Run():End");
compact_->status = status;
Parallelize L0-L1 Compaction: Restructure Compaction Job Summary: As of now compactions involving files from Level 0 and Level 1 are single threaded because the files in L0, although sorted, are not range partitioned like the other levels. This means that during L0-L1 compaction each file from L1 needs to be merged with potentially all the files from L0. This attempt to parallelize the L0-L1 compaction assigns a thread and a corresponding iterator to each L1 file that then considers only the key range found in that L1 file and only the L0 files that have those keys (and only the specific portion of those L0 files in which those keys are found). In this way the overlap is minimized and potentially eliminated between different iterators focusing on the same files. The first step is to restructure the compaction logic to break L0-L1 compactions into multiple, smaller, sequential compactions. Eventually each of these smaller jobs will be run simultaneously. Areas to pay extra attention to are # Correct aggregation of compaction job statistics across multiple threads # Proper opening/closing of output files (make sure each thread's is unique) # Keys that span multiple L1 files # Skewed distributions of keys within L0 files Test Plan: Make and run db_test (newer version has separate compaction tests) and compaction_job_stats_test Reviewers: igor, noetzli, anthony, sdong, yhchiang Reviewed By: yhchiang Subscribers: MarkCallaghan, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D42699
10 years ago
return status;
}
Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
assert(compact_);
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_INSTALL);
db_mutex_->AssertHeld();
Status status = compact_->status;
ColumnFamilyData* cfd = compact_->compaction->column_family_data();
assert(cfd);
int output_level = compact_->compaction->output_level();
cfd->internal_stats()->AddCompactionStats(output_level, thread_pri_,
compaction_stats_);
if (status.ok()) {
status = InstallCompactionResults(mutable_cf_options);
}
if (!versions_->io_status().ok()) {
io_status_ = versions_->io_status();
}
VersionStorageInfo::LevelSummaryStorage tmp;
auto vstorage = cfd->current()->storage_info();
const auto& stats = compaction_stats_.stats;
double read_write_amp = 0.0;
double write_amp = 0.0;
double bytes_read_per_sec = 0;
double bytes_written_per_sec = 0;
const uint64_t bytes_read_non_output_and_blob =
stats.bytes_read_non_output_levels + stats.bytes_read_blob;
const uint64_t bytes_read_all =
stats.bytes_read_output_level + bytes_read_non_output_and_blob;
const uint64_t bytes_written_all =
stats.bytes_written + stats.bytes_written_blob;
if (bytes_read_non_output_and_blob > 0) {
read_write_amp = (bytes_written_all + bytes_read_all) /
static_cast<double>(bytes_read_non_output_and_blob);
write_amp =
bytes_written_all / static_cast<double>(bytes_read_non_output_and_blob);
}
if (stats.micros > 0) {
bytes_read_per_sec = bytes_read_all / static_cast<double>(stats.micros);
bytes_written_per_sec =
bytes_written_all / static_cast<double>(stats.micros);
}
const std::string& column_family_name = cfd->GetName();
constexpr double kMB = 1048576.0;
ROCKS_LOG_BUFFER(
log_buffer_,
"[%s] compacted to: %s, MB/sec: %.1f rd, %.1f wr, level %d, "
"files in(%d, %d) out(%d +%d blob) "
"MB in(%.1f, %.1f +%.1f blob) out(%.1f +%.1f blob), "
"read-write-amplify(%.1f) write-amplify(%.1f) %s, records in: %" PRIu64
", records dropped: %" PRIu64 " output_compression: %s\n",
column_family_name.c_str(), vstorage->LevelSummary(&tmp),
bytes_read_per_sec, bytes_written_per_sec,
compact_->compaction->output_level(),
stats.num_input_files_in_non_output_levels,
stats.num_input_files_in_output_level, stats.num_output_files,
stats.num_output_files_blob, stats.bytes_read_non_output_levels / kMB,
stats.bytes_read_output_level / kMB, stats.bytes_read_blob / kMB,
stats.bytes_written / kMB, stats.bytes_written_blob / kMB, read_write_amp,
write_amp, status.ToString().c_str(), stats.num_input_records,
stats.num_dropped_records,
CompressionTypeToString(compact_->compaction->output_compression())
.c_str());
const auto& blob_files = vstorage->GetBlobFiles();
if (!blob_files.empty()) {
Use a sorted vector instead of a map to store blob file metadata (#9526) Summary: The patch replaces `std::map` with a sorted `std::vector` for `VersionStorageInfo::blob_files_` and preallocates the space for the `vector` before saving the `BlobFileMetaData` into the new `VersionStorageInfo` in `VersionBuilder::Rep::SaveBlobFilesTo`. These changes reduce the time the DB mutex is held while saving new `Version`s, and using a sorted `vector` also makes lookups faster thanks to better memory locality. In addition, the patch introduces helper methods `VersionStorageInfo::GetBlobFileMetaData` and `VersionStorageInfo::GetBlobFileMetaDataLB` that can be used by clients to perform lookups in the `vector`, and does some general cleanup in the parts of code where blob file metadata are used. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9526 Test Plan: Ran `make check` and the crash test script for a while. Performance was tested using a load-optimized benchmark (`fillseq` with vector memtable, no WAL) and small file sizes so that a significant number of files are produced: ``` numactl --interleave=all ./db_bench --benchmarks=fillseq --allow_concurrent_memtable_write=false --level0_file_num_compaction_trigger=4 --level0_slowdown_writes_trigger=20 --level0_stop_writes_trigger=30 --max_background_jobs=8 --max_write_buffer_number=8 --db=/data/ltamasi-dbbench --wal_dir=/data/ltamasi-dbbench --num=800000000 --num_levels=8 --key_size=20 --value_size=400 --block_size=8192 --cache_size=51539607552 --cache_numshardbits=6 --compression_max_dict_bytes=0 --compression_ratio=0.5 --compression_type=lz4 --bytes_per_sync=8388608 --cache_index_and_filter_blocks=1 --cache_high_pri_pool_ratio=0.5 --benchmark_write_rate_limit=0 --write_buffer_size=16777216 --target_file_size_base=16777216 --max_bytes_for_level_base=67108864 --verify_checksum=1 --delete_obsolete_files_period_micros=62914560 --max_bytes_for_level_multiplier=8 --statistics=0 --stats_per_interval=1 --stats_interval_seconds=20 --histogram=1 --memtablerep=skip_list --bloom_bits=10 --open_files=-1 --subcompactions=1 --compaction_style=0 --min_level_to_compress=3 --level_compaction_dynamic_level_bytes=true --pin_l0_filter_and_index_blocks_in_cache=1 --soft_pending_compaction_bytes_limit=167503724544 --hard_pending_compaction_bytes_limit=335007449088 --min_level_to_compress=0 --use_existing_db=0 --sync=0 --threads=1 --memtablerep=vector --allow_concurrent_memtable_write=false --disable_wal=1 --enable_blob_files=1 --blob_file_size=16777216 --min_blob_size=0 --blob_compression_type=lz4 --enable_blob_garbage_collection=1 --seed=<some value> ``` Final statistics before the patch: ``` Cumulative writes: 0 writes, 700M keys, 0 commit groups, 0.0 writes per commit group, ingest: 284.62 GB, 121.27 MB/s Interval writes: 0 writes, 334K keys, 0 commit groups, 0.0 writes per commit group, ingest: 139.28 MB, 72.46 MB/s ``` With the patch: ``` Cumulative writes: 0 writes, 760M keys, 0 commit groups, 0.0 writes per commit group, ingest: 308.66 GB, 131.52 MB/s Interval writes: 0 writes, 445K keys, 0 commit groups, 0.0 writes per commit group, ingest: 185.35 MB, 93.15 MB/s ``` Total time to complete the benchmark is 2611 seconds with the patch, down from 2986 secs. Reviewed By: riversand963 Differential Revision: D34082728 Pulled By: ltamasi fbshipit-source-id: fc598abf676dce436734d06bb9d2d99a26a004fc
3 years ago
assert(blob_files.front());
assert(blob_files.back());
ROCKS_LOG_BUFFER(
log_buffer_,
"[%s] Blob file summary: head=%" PRIu64 ", tail=%" PRIu64 "\n",
column_family_name.c_str(), blob_files.front()->GetBlobFileNumber(),
blob_files.back()->GetBlobFileNumber());
}
if (compaction_stats_.has_penultimate_level_output) {
ROCKS_LOG_BUFFER(
log_buffer_,
"[%s] has Penultimate Level output: %" PRIu64
", level %d, number of files: %" PRIu64 ", number of records: %" PRIu64,
column_family_name.c_str(),
compaction_stats_.penultimate_level_stats.bytes_written,
compact_->compaction->GetPenultimateLevel(),
compaction_stats_.penultimate_level_stats.num_output_files,
compaction_stats_.penultimate_level_stats.num_output_records);
}
UpdateCompactionJobStats(stats);
auto stream = event_logger_->LogToBuffer(log_buffer_, 8192);
stream << "job" << job_id_ << "event"
<< "compaction_finished"
<< "compaction_time_micros" << stats.micros
<< "compaction_time_cpu_micros" << stats.cpu_micros << "output_level"
<< compact_->compaction->output_level() << "num_output_files"
<< stats.num_output_files << "total_output_size"
<< stats.bytes_written;
if (stats.num_output_files_blob > 0) {
stream << "num_blob_output_files" << stats.num_output_files_blob
<< "total_blob_output_size" << stats.bytes_written_blob;
}
stream << "num_input_records" << stats.num_input_records
<< "num_output_records" << stats.num_output_records
<< "num_subcompactions" << compact_->sub_compact_states.size()
<< "output_compression"
<< CompressionTypeToString(compact_->compaction->output_compression());
stream << "num_single_delete_mismatches"
<< compaction_job_stats_->num_single_del_mismatch;
stream << "num_single_delete_fallthrough"
<< compaction_job_stats_->num_single_del_fallthru;
if (measure_io_stats_) {
stream << "file_write_nanos" << compaction_job_stats_->file_write_nanos;
stream << "file_range_sync_nanos"
<< compaction_job_stats_->file_range_sync_nanos;
stream << "file_fsync_nanos" << compaction_job_stats_->file_fsync_nanos;
stream << "file_prepare_write_nanos"
<< compaction_job_stats_->file_prepare_write_nanos;
}
stream << "lsm_state";
stream.StartArray();
for (int level = 0; level < vstorage->num_levels(); ++level) {
stream << vstorage->NumLevelFiles(level);
}
stream.EndArray();
if (!blob_files.empty()) {
Use a sorted vector instead of a map to store blob file metadata (#9526) Summary: The patch replaces `std::map` with a sorted `std::vector` for `VersionStorageInfo::blob_files_` and preallocates the space for the `vector` before saving the `BlobFileMetaData` into the new `VersionStorageInfo` in `VersionBuilder::Rep::SaveBlobFilesTo`. These changes reduce the time the DB mutex is held while saving new `Version`s, and using a sorted `vector` also makes lookups faster thanks to better memory locality. In addition, the patch introduces helper methods `VersionStorageInfo::GetBlobFileMetaData` and `VersionStorageInfo::GetBlobFileMetaDataLB` that can be used by clients to perform lookups in the `vector`, and does some general cleanup in the parts of code where blob file metadata are used. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9526 Test Plan: Ran `make check` and the crash test script for a while. Performance was tested using a load-optimized benchmark (`fillseq` with vector memtable, no WAL) and small file sizes so that a significant number of files are produced: ``` numactl --interleave=all ./db_bench --benchmarks=fillseq --allow_concurrent_memtable_write=false --level0_file_num_compaction_trigger=4 --level0_slowdown_writes_trigger=20 --level0_stop_writes_trigger=30 --max_background_jobs=8 --max_write_buffer_number=8 --db=/data/ltamasi-dbbench --wal_dir=/data/ltamasi-dbbench --num=800000000 --num_levels=8 --key_size=20 --value_size=400 --block_size=8192 --cache_size=51539607552 --cache_numshardbits=6 --compression_max_dict_bytes=0 --compression_ratio=0.5 --compression_type=lz4 --bytes_per_sync=8388608 --cache_index_and_filter_blocks=1 --cache_high_pri_pool_ratio=0.5 --benchmark_write_rate_limit=0 --write_buffer_size=16777216 --target_file_size_base=16777216 --max_bytes_for_level_base=67108864 --verify_checksum=1 --delete_obsolete_files_period_micros=62914560 --max_bytes_for_level_multiplier=8 --statistics=0 --stats_per_interval=1 --stats_interval_seconds=20 --histogram=1 --memtablerep=skip_list --bloom_bits=10 --open_files=-1 --subcompactions=1 --compaction_style=0 --min_level_to_compress=3 --level_compaction_dynamic_level_bytes=true --pin_l0_filter_and_index_blocks_in_cache=1 --soft_pending_compaction_bytes_limit=167503724544 --hard_pending_compaction_bytes_limit=335007449088 --min_level_to_compress=0 --use_existing_db=0 --sync=0 --threads=1 --memtablerep=vector --allow_concurrent_memtable_write=false --disable_wal=1 --enable_blob_files=1 --blob_file_size=16777216 --min_blob_size=0 --blob_compression_type=lz4 --enable_blob_garbage_collection=1 --seed=<some value> ``` Final statistics before the patch: ``` Cumulative writes: 0 writes, 700M keys, 0 commit groups, 0.0 writes per commit group, ingest: 284.62 GB, 121.27 MB/s Interval writes: 0 writes, 334K keys, 0 commit groups, 0.0 writes per commit group, ingest: 139.28 MB, 72.46 MB/s ``` With the patch: ``` Cumulative writes: 0 writes, 760M keys, 0 commit groups, 0.0 writes per commit group, ingest: 308.66 GB, 131.52 MB/s Interval writes: 0 writes, 445K keys, 0 commit groups, 0.0 writes per commit group, ingest: 185.35 MB, 93.15 MB/s ``` Total time to complete the benchmark is 2611 seconds with the patch, down from 2986 secs. Reviewed By: riversand963 Differential Revision: D34082728 Pulled By: ltamasi fbshipit-source-id: fc598abf676dce436734d06bb9d2d99a26a004fc
3 years ago
assert(blob_files.front());
stream << "blob_file_head" << blob_files.front()->GetBlobFileNumber();
assert(blob_files.back());
stream << "blob_file_tail" << blob_files.back()->GetBlobFileNumber();
}
if (compaction_stats_.has_penultimate_level_output) {
InternalStats::CompactionStats& pl_stats =
compaction_stats_.penultimate_level_stats;
stream << "penultimate_level_num_output_files" << pl_stats.num_output_files;
stream << "penultimate_level_bytes_written" << pl_stats.bytes_written;
stream << "penultimate_level_num_output_records"
<< pl_stats.num_output_records;
stream << "penultimate_level_num_output_files_blob"
<< pl_stats.num_output_files_blob;
stream << "penultimate_level_bytes_written_blob"
<< pl_stats.bytes_written_blob;
}
CleanupCompaction();
return status;
}
void CompactionJob::NotifyOnSubcompactionBegin(
SubcompactionState* sub_compact) {
Compaction* c = compact_->compaction;
if (db_options_.listeners.empty()) {
return;
}
if (shutting_down_->load(std::memory_order_acquire)) {
return;
}
if (c->is_manual_compaction() &&
manual_compaction_canceled_.load(std::memory_order_acquire)) {
return;
}
sub_compact->notify_on_subcompaction_completion = true;
SubcompactionJobInfo info{};
sub_compact->BuildSubcompactionJobInfo(info);
info.job_id = static_cast<int>(job_id_);
info.thread_id = env_->GetThreadID();
for (const auto& listener : db_options_.listeners) {
listener->OnSubcompactionBegin(info);
}
info.status.PermitUncheckedError();
}
void CompactionJob::NotifyOnSubcompactionCompleted(
SubcompactionState* sub_compact) {
if (db_options_.listeners.empty()) {
return;
}
if (shutting_down_->load(std::memory_order_acquire)) {
return;
}
if (sub_compact->notify_on_subcompaction_completion == false) {
return;
}
SubcompactionJobInfo info{};
sub_compact->BuildSubcompactionJobInfo(info);
info.job_id = static_cast<int>(job_id_);
info.thread_id = env_->GetThreadID();
for (const auto& listener : db_options_.listeners) {
listener->OnSubcompactionCompleted(info);
}
}
void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
assert(sub_compact);
assert(sub_compact->compaction);
if (db_options_.compaction_service) {
CompactionServiceJobStatus comp_status =
ProcessKeyValueCompactionWithCompactionService(sub_compact);
if (comp_status == CompactionServiceJobStatus::kSuccess ||
comp_status == CompactionServiceJobStatus::kFailure) {
return;
}
// fallback to local compaction
assert(comp_status == CompactionServiceJobStatus::kUseLocal);
}
uint64_t prev_cpu_micros = db_options_.clock->CPUMicros();
Compaction Support for Range Deletion Summary: This diff introduces RangeDelAggregator, which takes ownership of iterators provided to it via AddTombstones(). The tombstones are organized in a two-level map (snapshot stripe -> begin key -> tombstone). Tombstone creation avoids data copy by holding Slices returned by the iterator, which remain valid thanks to pinning. For compaction, we create a hierarchical range tombstone iterator with structure matching the iterator over compaction input data. An aggregator based on that iterator is used by CompactionIterator to determine which keys are covered by range tombstones. In case of merge operand, the same aggregator is used by MergeHelper. Upon finishing each file in the compaction, relevant range tombstones are added to the output file's range tombstone metablock and file boundaries are updated accordingly. To check whether a key is covered by range tombstone, RangeDelAggregator::ShouldDelete() considers tombstones in the key's snapshot stripe. When this function is used outside of compaction, it also checks newer stripes, which can contain covering tombstones. Currently the intra-stripe check involves a linear scan; however, in the future we plan to collapse ranges within a stripe such that binary search can be used. RangeDelAggregator::AddToBuilder() adds all range tombstones in the table's key-range to a new table's range tombstone meta-block. Since range tombstones may fall in the gap between files, we may need to extend some files' key-ranges. The strategy is (1) first file extends as far left as possible and other files do not extend left, (2) all files extend right until either the start of the next file or the end of the last range tombstone in the gap, whichever comes first. One other notable change is adding release/move semantics to ScopedArenaIterator such that it can be used to transfer ownership of an arena-allocated iterator, similar to how unique_ptr is used for malloc'd data. Depends on D61473 Test Plan: compaction_iterator_test, mock_table, end-to-end tests in D63927 Reviewers: sdong, IslamAbdelRahman, wanning, yhchiang, lightmark Reviewed By: lightmark Subscribers: andrewkr, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D62205
8 years ago
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
// Create compaction filter and fail the compaction if
// IgnoreSnapshots() = false because it is not supported anymore
const CompactionFilter* compaction_filter =
cfd->ioptions()->compaction_filter;
std::unique_ptr<CompactionFilter> compaction_filter_from_factory = nullptr;
if (compaction_filter == nullptr) {
compaction_filter_from_factory =
sub_compact->compaction->CreateCompactionFilter();
compaction_filter = compaction_filter_from_factory.get();
}
if (compaction_filter != nullptr && !compaction_filter->IgnoreSnapshots()) {
sub_compact->status = Status::NotSupported(
"CompactionFilter::IgnoreSnapshots() = false is not supported "
"anymore.");
return;
}
NotifyOnSubcompactionBegin(sub_compact);
auto range_del_agg = std::make_unique<CompactionRangeDelAggregator>(
User-defined timestamp support for `DeleteRange()` (#10661) Summary: Add user-defined timestamp support for range deletion. The new API is `DeleteRange(opt, cf, begin_key, end_key, ts)`. Most of the change is to update the comparator to compare without timestamp. Other than that, major changes are - internal range tombstone data structures (`FragmentedRangeTombstoneList`, `RangeTombstone`, etc.) to store timestamps. - Garbage collection of range tombstones and range tombstone covered keys during compaction. - Get()/MultiGet() to return the timestamp of a range tombstone when needed. - Get/Iterator with range tombstones bounded by readoptions.timestamp. - timestamp crash test now issues DeleteRange by default. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10661 Test Plan: - Added unit test: `make check` - Stress test: `python3 tools/db_crashtest.py --enable_ts whitebox --readpercent=57 --prefixpercent=4 --writepercent=25 -delpercent=5 --iterpercent=5 --delrangepercent=4` - Ran `db_bench` to measure regression when timestamp is not enabled. The tests are for write (with some range deletion) and iterate with DB fitting in memory: `./db_bench--benchmarks=fillrandom,seekrandom --writes_per_range_tombstone=200 --max_write_buffer_number=100 --min_write_buffer_number_to_merge=100 --writes=500000 --reads=500000 --seek_nexts=10 --disable_auto_compactions -disable_wal=true --max_num_range_tombstones=1000`. Did not see consistent regression in no timestamp case. | micros/op | fillrandom | seekrandom | | --- | --- | --- | |main| 2.58 |10.96| |PR 10661| 2.68 |10.63| Reviewed By: riversand963 Differential Revision: D39441192 Pulled By: cbi42 fbshipit-source-id: f05aca3c41605caf110daf0ff405919f300ddec2
2 years ago
&cfd->internal_comparator(), existing_snapshots_, &full_history_ts_low_,
&trim_ts_);
// TODO: since we already use C++17, should use
// std::optional<const Slice> instead.
const std::optional<Slice> start = sub_compact->start;
const std::optional<Slice> end = sub_compact->end;
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
std::optional<Slice> start_without_ts;
std::optional<Slice> end_without_ts;
ReadOptions read_options;
read_options.verify_checksums = true;
read_options.fill_cache = false;
Set Read rate limiter priority dynamically and pass it to FS (#9996) Summary: ### Context: Background compactions and flush generate large reads and writes, and can be long running, especially for universal compaction. In some cases, this can impact foreground reads and writes by users. ### Solution User, Flush, and Compaction reads share some code path. For this task, we update the rate_limiter_priority in ReadOptions for code paths (e.g. FindTable (mainly in BlockBasedTable::Open()) and various iterators), and eventually update the rate_limiter_priority in IOOptions for FSRandomAccessFile. **This PR is for the Read path.** The **Read:** dynamic priority for different state are listed as follows: | State | Normal | Delayed | Stalled | | ----- | ------ | ------- | ------- | | Flush (verification read in BuildTable()) | IO_USER | IO_USER | IO_USER | | Compaction | IO_LOW | IO_USER | IO_USER | | User | User provided | User provided | User provided | We will respect the read_options that the user provided and will not set it. The only sst read for Flush is the verification read in BuildTable(). It claims to be "regard as user read". **Details** 1. Set read_options.rate_limiter_priority dynamically: - User: Do not update the read_options. Use the read_options that the user provided. - Compaction: Update read_options in CompactionJob::ProcessKeyValueCompaction(). - Flush: Update read_options in BuildTable(). 2. Pass the rate limiter priority to FSRandomAccessFile functions: - After calling the FindTable(), read_options is passed through GetTableReader(table_cache.cc), BlockBasedTableFactory::NewTableReader(block_based_table_factory.cc), and BlockBasedTable::Open(). The Open() needs some updates for the ReadOptions variable and the updates are also needed for the called functions, including PrefetchTail(), PrepareIOOptions(), ReadFooterFromFile(), ReadMetaIndexblock(), ReadPropertiesBlock(), PrefetchIndexAndFilterBlocks(), and ReadRangeDelBlock(). - In RandomAccessFileReader, the functions to be updated include Read(), MultiRead(), ReadAsync(), and Prefetch(). - Update the downstream functions of NewIndexIterator(), NewDataBlockIterator(), and BlockBasedTableIterator(). ### Test Plans Add unit tests. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9996 Reviewed By: anand1976 Differential Revision: D36452483 Pulled By: gitbw95 fbshipit-source-id: 60978204a4f849bb9261cb78d9bc1cb56d6008cf
3 years ago
read_options.rate_limiter_priority = GetRateLimiterPriority();
// Compaction iterators shouldn't be confined to a single prefix.
// Compactions use Seek() for
// (a) concurrent compactions,
// (b) CompactionFilter::Decision::kRemoveAndSkipUntil.
read_options.total_order_seek = true;
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
// Remove the timestamps from boundaries because boundaries created in
// GenSubcompactionBoundaries doesn't strip away the timestamp.
size_t ts_sz = cfd->user_comparator()->timestamp_size();
if (start.has_value()) {
read_options.iterate_lower_bound = &start.value();
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
if (ts_sz > 0) {
start_without_ts = StripTimestampFromUserKey(start.value(), ts_sz);
read_options.iterate_lower_bound = &start_without_ts.value();
}
}
if (end.has_value()) {
read_options.iterate_upper_bound = &end.value();
Provide support for subcompactions with user-defined timestamps (#10344) Summary: The subcompaction logic currently picks file boundaries as subcompaction boundaries. This is not compatible with user-defined timestamps because of two issues. Issue1: ReadOptions.iterate_lower_bound and ReadOptions.iterate_upper_bound contains timestamps which results in assertion failure as BlockBasedTableIterator expects bounds to be without timestamps. As result, because of wrong comparison end key is returned as user_key resulting in assertion failure. Issue2: Since it might result in two keys that only differ by user timestamp getting processed by two different subcompactions (and thus two different CompactionIterator state machines), which in turn can cause data correction issues. This PR provide support to reenable subcompactions with user-defined timestamps. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10344 Test Plan: Added new unit test - Without fix for Issue1 unit test MultipleSubCompactions fails with error: ``` db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterat│ or::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Received signal 6 (Aborted) │ #0 /usr/local/fbcode/platform009/lib/libc.so.6(gsignal+0x100) [0x7f8fbbbfe530] db_with_timestamp_compaction_test: ./db/compaction/clipping_iterator.h:247: void rocksdb::ClippingIterator::AssertBounds(): Assertion `!valid_ || !end_ || cmp_->Compare(key(), *end_) < 0' failed. Aborted (core dumped) ``` Ran stress test `make crash_test_with_ts -j32` Reviewed By: riversand963 Differential Revision: D38220841 Pulled By: akankshamahajan15 fbshipit-source-id: 5d5cae2bd37fcaeba1e77fce0a69070ad4158ccb
2 years ago
if (ts_sz > 0) {
end_without_ts = StripTimestampFromUserKey(end.value(), ts_sz);
read_options.iterate_upper_bound = &end_without_ts.value();
}
}
// Although the v2 aggregator is what the level iterator(s) know about,
// the AddTombstones calls will be propagated down to the v1 aggregator.
std::unique_ptr<InternalIterator> raw_input(versions_->MakeInputIterator(
read_options, sub_compact->compaction, range_del_agg.get(),
file_options_for_read_, start, end));
InternalIterator* input = raw_input.get();
IterKey start_ikey;
IterKey end_ikey;
Slice start_slice;
Slice end_slice;
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
static constexpr char kMaxTs[] =
"\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff";
Slice ts_slice;
std::string max_ts;
if (ts_sz > 0) {
if (ts_sz <= strlen(kMaxTs)) {
ts_slice = Slice(kMaxTs, ts_sz);
} else {
max_ts = std::string(ts_sz, '\xff');
ts_slice = Slice(max_ts);
}
}
if (start.has_value()) {
start_ikey.SetInternalKey(start.value(), kMaxSequenceNumber,
kValueTypeForSeek);
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
if (ts_sz > 0) {
start_ikey.UpdateInternalKey(kMaxSequenceNumber, kValueTypeForSeek,
&ts_slice);
}
start_slice = start_ikey.GetInternalKey();
}
if (end.has_value()) {
end_ikey.SetInternalKey(end.value(), kMaxSequenceNumber, kValueTypeForSeek);
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
if (ts_sz > 0) {
end_ikey.UpdateInternalKey(kMaxSequenceNumber, kValueTypeForSeek,
&ts_slice);
}
end_slice = end_ikey.GetInternalKey();
}
std::unique_ptr<InternalIterator> clip;
if (start.has_value() || end.has_value()) {
clip = std::make_unique<ClippingIterator>(
raw_input.get(), start.has_value() ? &start_slice : nullptr,
end.has_value() ? &end_slice : nullptr, &cfd->internal_comparator());
input = clip.get();
}
std::unique_ptr<InternalIterator> blob_counter;
if (sub_compact->compaction->DoesInputReferenceBlobFiles()) {
BlobGarbageMeter* meter = sub_compact->Current().CreateBlobGarbageMeter();
blob_counter = std::make_unique<BlobCountingIterator>(input, meter);
input = blob_counter.get();
}
std::unique_ptr<InternalIterator> trim_history_iter;
Fix a bug by setting up subcompaction bounds properly (#10658) Summary: When user-defined timestamp is enabled, subcompaction bounds should be set up properly. When creating InputIterator for the compaction, the `start` and `end` should have their timestamp portions set to kMaxTimestamp, which is the highest possible timestamp. This is similar to what we do with setting up their sequence numbers to `kMaxSequenceNumber`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10658 Test Plan: ```bash make check rm -rf /dev/shm/rocksdb/* && mkdir /dev/shm/rocksdb/rocksdb_crashtest_expected && ./db_stress --allow_data_in_errors=True --clear_column_family_one_in=0 --continuous_verification_interval=0 --data_block_index_type=1 --db=/dev/shm/rocksdb//rocksdb_crashtest_blackbox --delpercent=5 --delrangepercent=0 --expected_values_dir=/dev/shm/rocksdb//rocksdb_crashtest_expected --iterpercent=0 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=25000000 --max_write_batch_group_size_bytes=1048576 --nooverwritepercent=1 --ops_per_thread=300000 --paranoid_file_checks=1 --partition_filters=0 --prefix_size=8 --prefixpercent=5 --readpercent=30 --reopen=0 --snapshot_hold_ops=100000 --subcompactions=4 --target_file_size_base=65536 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 --use_multiget=1 --user_timestamp_size=8 --value_size_mult=32 --verify_checksum=1 --write_buffer_size=65536 --writepercent=60 -disable_wal=1 -column_families=1 ``` Reviewed By: akankshamahajan15 Differential Revision: D39393402 Pulled By: riversand963 fbshipit-source-id: f276e35b19fce51a175c368a502fb0718d1f3871
2 years ago
if (ts_sz > 0 && !trim_ts_.empty()) {
trim_history_iter = std::make_unique<HistoryTrimmingIterator>(
input, cfd->user_comparator(), trim_ts_);
input = trim_history_iter.get();
}
input->SeekToFirst();
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_PROCESS_KV);
// I/O measurement variables
PerfLevel prev_perf_level = PerfLevel::kEnableTime;
const uint64_t kRecordStatsEvery = 1000;
uint64_t prev_write_nanos = 0;
uint64_t prev_fsync_nanos = 0;
uint64_t prev_range_sync_nanos = 0;
uint64_t prev_prepare_write_nanos = 0;
uint64_t prev_cpu_write_nanos = 0;
uint64_t prev_cpu_read_nanos = 0;
if (measure_io_stats_) {
prev_perf_level = GetPerfLevel();
SetPerfLevel(PerfLevel::kEnableTimeAndCPUTimeExceptForMutex);
prev_write_nanos = IOSTATS(write_nanos);
prev_fsync_nanos = IOSTATS(fsync_nanos);
prev_range_sync_nanos = IOSTATS(range_sync_nanos);
prev_prepare_write_nanos = IOSTATS(prepare_write_nanos);
prev_cpu_write_nanos = IOSTATS(cpu_write_nanos);
prev_cpu_read_nanos = IOSTATS(cpu_read_nanos);
}
MergeHelper merge(
env_, cfd->user_comparator(), cfd->ioptions()->merge_operator.get(),
compaction_filter, db_options_.info_log.get(),
false /* internal key corruption is expected */,
existing_snapshots_.empty() ? 0 : existing_snapshots_.back(),
snapshot_checker_, compact_->compaction->level(), db_options_.stats);
const MutableCFOptions* mutable_cf_options =
sub_compact->compaction->mutable_cf_options();
assert(mutable_cf_options);
std::vector<std::string> blob_file_paths;
// TODO: BlobDB to support output_to_penultimate_level compaction, which needs
// 2 builders, so may need to move to `CompactionOutputs`
std::unique_ptr<BlobFileBuilder> blob_file_builder(
Make it possible to enable blob files starting from a certain LSM tree level (#10077) Summary: Currently, if blob files are enabled (i.e. `enable_blob_files` is true), large values are extracted both during flush/recovery (when SST files are written into level 0 of the LSM tree) and during compaction into any LSM tree level. For certain use cases that have a mix of short-lived and long-lived values, it might make sense to support extracting large values only during compactions whose output level is greater than or equal to a specified LSM tree level (e.g. compactions into L1/L2/... or above). This could reduce the space amplification caused by large values that are turned into garbage shortly after being written at the price of some write amplification incurred by long-lived values whose extraction to blob files is delayed. In order to achieve this, we would like to do the following: - Add a new configuration option `blob_file_starting_level` (default: 0) to `AdvancedColumnFamilyOptions` (and `MutableCFOptions` and extend the related logic) - Instantiate `BlobFileBuilder` in `BuildTable` (used during flush and recovery, where the LSM tree level is L0) and `CompactionJob` iff `enable_blob_files` is set and the LSM tree level is `>= blob_file_starting_level` - Add unit tests for the new functionality, and add the new option to our stress tests (`db_stress` and `db_crashtest.py` ) - Add the new option to our benchmarking tool `db_bench` and the BlobDB benchmark script `run_blob_bench.sh` - Add the new option to the `ldb` tool (see https://github.com/facebook/rocksdb/wiki/Administration-and-Data-Access-Tool) - Ideally extend the C and Java bindings with the new option - Update the BlobDB wiki to document the new option. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10077 Reviewed By: ltamasi Differential Revision: D36884156 Pulled By: gangliao fbshipit-source-id: 942bab025f04633edca8564ed64791cb5e31627d
3 years ago
(mutable_cf_options->enable_blob_files &&
sub_compact->compaction->output_level() >=
mutable_cf_options->blob_file_starting_level)
? new BlobFileBuilder(
versions_, fs_.get(),
sub_compact->compaction->immutable_options(),
mutable_cf_options, &file_options_, db_id_, db_session_id_,
job_id_, cfd->GetID(), cfd->GetName(), Env::IOPriority::IO_LOW,
write_hint_, io_tracer_, blob_callback_,
BlobFileCreationReason::kCompaction, &blob_file_paths,
sub_compact->Current().GetBlobFileAdditionsPtr())
: nullptr);
TEST_SYNC_POINT("CompactionJob::Run():Inprogress");
TEST_SYNC_POINT_CALLBACK(
"CompactionJob::Run():PausingManualCompaction:1",
reinterpret_cast<void*>(
const_cast<std::atomic<bool>*>(&manual_compaction_canceled_)));
const std::string* const full_history_ts_low =
full_history_ts_low_.empty() ? nullptr : &full_history_ts_low_;
CompactionIterator sees consistent view of which keys are committed (#9830) Summary: **This PR does not affect the functionality of `DB` and write-committed transactions.** `CompactionIterator` uses `KeyCommitted(seq)` to determine if a key in the database is committed. As the name 'write-committed' implies, if write-committed policy is used, a key exists in the database only if it is committed. In fact, the implementation of `KeyCommitted()` is as follows: ``` inline bool KeyCommitted(SequenceNumber seq) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(seq, kMaxSequence) == SnapshotCheckerResult::kInSnapshot; } ``` With that being said, we focus on write-prepared/write-unprepared transactions. A few notes: - A key can exist in the db even if it's uncommitted. Therefore, we rely on `snapshot_checker_` to determine data visibility. We also require that all writes go through transaction API instead of the raw `WriteBatch` + `Write`, thus at most one uncommitted version of one user key can exist in the database. - `CompactionIterator` outputs a key as long as the key is uncommitted. Due to the above reasons, it is possible that `CompactionIterator` decides to output an uncommitted key without doing further checks on the key (`NextFromInput()`). By the time the key is being prepared for output, the key becomes committed because the `snapshot_checker_(seq, kMaxSequence)` becomes true in the implementation of `KeyCommitted()`. Then `CompactionIterator` will try to zero its sequence number and hit assertion error if the key is a tombstone. To fix this issue, we should make the `CompactionIterator` see a consistent view of the input keys. Note that for write-prepared/write-unprepared, the background flush/compaction jobs already take a "job snapshot" before starting processing keys. The job snapshot is released only after the entire flush/compaction finishes. We can use this snapshot to determine whether a key is committed or not with minor change to `KeyCommitted()`. ``` inline bool KeyCommitted(SequenceNumber sequence) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(sequence, job_snapshot_) == SnapshotCheckerResult::kInSnapshot; } ``` As a result, whether a key is committed or not will remain a constant throughout compaction, causing no trouble for `CompactionIterator`s assertions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9830 Test Plan: make check Reviewed By: ltamasi Differential Revision: D35561162 Pulled By: riversand963 fbshipit-source-id: 0e00d200c195240341cfe6d34cbc86798b315b9f
3 years ago
const SequenceNumber job_snapshot_seq =
job_context_ ? job_context_->GetJobSnapshotSequence()
: kMaxSequenceNumber;
auto c_iter = std::make_unique<CompactionIterator>(
input, cfd->user_comparator(), &merge, versions_->LastSequence(),
CompactionIterator sees consistent view of which keys are committed (#9830) Summary: **This PR does not affect the functionality of `DB` and write-committed transactions.** `CompactionIterator` uses `KeyCommitted(seq)` to determine if a key in the database is committed. As the name 'write-committed' implies, if write-committed policy is used, a key exists in the database only if it is committed. In fact, the implementation of `KeyCommitted()` is as follows: ``` inline bool KeyCommitted(SequenceNumber seq) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(seq, kMaxSequence) == SnapshotCheckerResult::kInSnapshot; } ``` With that being said, we focus on write-prepared/write-unprepared transactions. A few notes: - A key can exist in the db even if it's uncommitted. Therefore, we rely on `snapshot_checker_` to determine data visibility. We also require that all writes go through transaction API instead of the raw `WriteBatch` + `Write`, thus at most one uncommitted version of one user key can exist in the database. - `CompactionIterator` outputs a key as long as the key is uncommitted. Due to the above reasons, it is possible that `CompactionIterator` decides to output an uncommitted key without doing further checks on the key (`NextFromInput()`). By the time the key is being prepared for output, the key becomes committed because the `snapshot_checker_(seq, kMaxSequence)` becomes true in the implementation of `KeyCommitted()`. Then `CompactionIterator` will try to zero its sequence number and hit assertion error if the key is a tombstone. To fix this issue, we should make the `CompactionIterator` see a consistent view of the input keys. Note that for write-prepared/write-unprepared, the background flush/compaction jobs already take a "job snapshot" before starting processing keys. The job snapshot is released only after the entire flush/compaction finishes. We can use this snapshot to determine whether a key is committed or not with minor change to `KeyCommitted()`. ``` inline bool KeyCommitted(SequenceNumber sequence) { // For non-txn-db and write-committed, snapshot_checker_ is always nullptr. return snapshot_checker_ == nullptr || snapshot_checker_->CheckInSnapshot(sequence, job_snapshot_) == SnapshotCheckerResult::kInSnapshot; } ``` As a result, whether a key is committed or not will remain a constant throughout compaction, causing no trouble for `CompactionIterator`s assertions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9830 Test Plan: make check Reviewed By: ltamasi Differential Revision: D35561162 Pulled By: riversand963 fbshipit-source-id: 0e00d200c195240341cfe6d34cbc86798b315b9f
3 years ago
&existing_snapshots_, earliest_write_conflict_snapshot_, job_snapshot_seq,
snapshot_checker_, env_, ShouldReportDetailedTime(env_, stats_),
/*expect_valid_internal_key=*/true, range_del_agg.get(),
blob_file_builder.get(), db_options_.allow_data_in_errors,
db_options_.enforce_single_del_contracts, manual_compaction_canceled_,
sub_compact->compaction, compaction_filter, shutting_down_,
db_options_.info_log, full_history_ts_low, preserve_time_min_seqno_,
preclude_last_level_min_seqno_);
c_iter->SeekToFirst();
// Assign range delete aggregator to the target output level, which makes sure
// it only output to single level
sub_compact->AssignRangeDelAggregator(std::move(range_del_agg));
const auto& c_iter_stats = c_iter->iter_stats();
// define the open and close functions for the compaction files, which will be
// used open/close output files when needed.
const CompactionFileOpenFunc open_file_func =
[this, sub_compact](CompactionOutputs& outputs) {
return this->OpenCompactionOutputFile(sub_compact, outputs);
};
const CompactionFileCloseFunc close_file_func =
[this, sub_compact](CompactionOutputs& outputs, const Status& status,
const Slice& next_table_min_key) {
return this->FinishCompactionOutputFile(status, sub_compact, outputs,
next_table_min_key);
};
Status status;
TEST_SYNC_POINT_CALLBACK(
"CompactionJob::ProcessKeyValueCompaction()::Processing",
reinterpret_cast<void*>(
const_cast<Compaction*>(sub_compact->compaction)));
while (status.ok() && !cfd->IsDropped() && c_iter->Valid()) {
// Invariant: c_iter.status() is guaranteed to be OK if c_iter->Valid()
// returns true.
assert(!end.has_value() || cfd->user_comparator()->Compare(
c_iter->user_key(), end.value()) < 0);
if (c_iter_stats.num_input_records % kRecordStatsEvery ==
kRecordStatsEvery - 1) {
RecordDroppedKeys(c_iter_stats, &sub_compact->compaction_job_stats);
c_iter->ResetRecordCounts();
RecordCompactionIOStats();
}
// Add current compaction_iterator key to target compaction output, if the
// output file needs to be close or open, it will call the `open_file_func`
// and `close_file_func`.
// TODO: it would be better to have the compaction file open/close moved
// into `CompactionOutputs` which has the output file information.
status = sub_compact->AddToOutput(*c_iter, open_file_func, close_file_func);
if (!status.ok()) {
break;
}
TEST_SYNC_POINT_CALLBACK(
"CompactionJob::Run():PausingManualCompaction:2",
reinterpret_cast<void*>(
const_cast<std::atomic<bool>*>(&manual_compaction_canceled_)));
c_iter->Next();
if (c_iter->status().IsManualCompactionPaused()) {
break;
}
}
sub_compact->compaction_job_stats.num_blobs_read =
c_iter_stats.num_blobs_read;
sub_compact->compaction_job_stats.total_blob_bytes_read =
c_iter_stats.total_blob_bytes_read;
sub_compact->compaction_job_stats.num_input_deletion_records =
c_iter_stats.num_input_deletion_records;
sub_compact->compaction_job_stats.num_corrupt_keys =
c_iter_stats.num_input_corrupt_records;
sub_compact->compaction_job_stats.num_single_del_fallthru =
c_iter_stats.num_single_del_fallthru;
sub_compact->compaction_job_stats.num_single_del_mismatch =
c_iter_stats.num_single_del_mismatch;
sub_compact->compaction_job_stats.total_input_raw_key_bytes +=
c_iter_stats.total_input_raw_key_bytes;
sub_compact->compaction_job_stats.total_input_raw_value_bytes +=
c_iter_stats.total_input_raw_value_bytes;
RecordTick(stats_, FILTER_OPERATION_TOTAL_TIME,
c_iter_stats.total_filter_time);
if (c_iter_stats.num_blobs_relocated > 0) {
RecordTick(stats_, BLOB_DB_GC_NUM_KEYS_RELOCATED,
c_iter_stats.num_blobs_relocated);
}
if (c_iter_stats.total_blob_bytes_relocated > 0) {
RecordTick(stats_, BLOB_DB_GC_BYTES_RELOCATED,
c_iter_stats.total_blob_bytes_relocated);
}
RecordDroppedKeys(c_iter_stats, &sub_compact->compaction_job_stats);
RecordCompactionIOStats();
if (status.ok() && cfd->IsDropped()) {
status =
Status::ColumnFamilyDropped("Column family dropped during compaction");
}
if ((status.ok() || status.IsColumnFamilyDropped()) &&
shutting_down_->load(std::memory_order_relaxed)) {
status = Status::ShutdownInProgress("Database shutdown");
}
if ((status.ok() || status.IsColumnFamilyDropped()) &&
(manual_compaction_canceled_.load(std::memory_order_relaxed))) {
status = Status::Incomplete(Status::SubCode::kManualCompactionPaused);
}
if (status.ok()) {
status = input->status();
}
if (status.ok()) {
status = c_iter->status();
}
// Call FinishCompactionOutputFile() even if status is not ok: it needs to
// close the output files. Open file function is also passed, in case there's
// only range-dels, no file was opened, to save the range-dels, it need to
// create a new output file.
status = sub_compact->CloseCompactionFiles(status, open_file_func,
close_file_func);
if (blob_file_builder) {
if (status.ok()) {
status = blob_file_builder->Finish();
} else {
blob_file_builder->Abandon(status);
}
blob_file_builder.reset();
sub_compact->Current().UpdateBlobStats();
}
sub_compact->compaction_job_stats.cpu_micros =
db_options_.clock->CPUMicros() - prev_cpu_micros;
if (measure_io_stats_) {
sub_compact->compaction_job_stats.file_write_nanos +=
IOSTATS(write_nanos) - prev_write_nanos;
sub_compact->compaction_job_stats.file_fsync_nanos +=
IOSTATS(fsync_nanos) - prev_fsync_nanos;
sub_compact->compaction_job_stats.file_range_sync_nanos +=
IOSTATS(range_sync_nanos) - prev_range_sync_nanos;
sub_compact->compaction_job_stats.file_prepare_write_nanos +=
IOSTATS(prepare_write_nanos) - prev_prepare_write_nanos;
sub_compact->compaction_job_stats.cpu_micros -=
(IOSTATS(cpu_write_nanos) - prev_cpu_write_nanos +
IOSTATS(cpu_read_nanos) - prev_cpu_read_nanos) /
1000;
if (prev_perf_level != PerfLevel::kEnableTimeAndCPUTimeExceptForMutex) {
SetPerfLevel(prev_perf_level);
}
}
#ifdef ROCKSDB_ASSERT_STATUS_CHECKED
if (!status.ok()) {
if (c_iter) {
c_iter->status().PermitUncheckedError();
}
if (input) {
input->status().PermitUncheckedError();
}
}
#endif // ROCKSDB_ASSERT_STATUS_CHECKED
blob_counter.reset();
clip.reset();
raw_input.reset();
sub_compact->status = status;
NotifyOnSubcompactionCompleted(sub_compact);
}
uint64_t CompactionJob::GetCompactionId(SubcompactionState* sub_compact) const {
return (uint64_t)job_id_ << 32 | sub_compact->sub_job_id;
}
void CompactionJob::RecordDroppedKeys(
const CompactionIterationStats& c_iter_stats,
CompactionJobStats* compaction_job_stats) {
if (c_iter_stats.num_record_drop_user > 0) {
RecordTick(stats_, COMPACTION_KEY_DROP_USER,
c_iter_stats.num_record_drop_user);
}
if (c_iter_stats.num_record_drop_hidden > 0) {
RecordTick(stats_, COMPACTION_KEY_DROP_NEWER_ENTRY,
c_iter_stats.num_record_drop_hidden);
if (compaction_job_stats) {
compaction_job_stats->num_records_replaced +=
c_iter_stats.num_record_drop_hidden;
}
}
if (c_iter_stats.num_record_drop_obsolete > 0) {
RecordTick(stats_, COMPACTION_KEY_DROP_OBSOLETE,
c_iter_stats.num_record_drop_obsolete);
if (compaction_job_stats) {
compaction_job_stats->num_expired_deletion_records +=
c_iter_stats.num_record_drop_obsolete;
}
}
if (c_iter_stats.num_record_drop_range_del > 0) {
RecordTick(stats_, COMPACTION_KEY_DROP_RANGE_DEL,
c_iter_stats.num_record_drop_range_del);
}
if (c_iter_stats.num_range_del_drop_obsolete > 0) {
RecordTick(stats_, COMPACTION_RANGE_DEL_DROP_OBSOLETE,
c_iter_stats.num_range_del_drop_obsolete);
}
if (c_iter_stats.num_optimized_del_drop_obsolete > 0) {
RecordTick(stats_, COMPACTION_OPTIMIZED_DEL_DROP_OBSOLETE,
c_iter_stats.num_optimized_del_drop_obsolete);
}
}
Status CompactionJob::FinishCompactionOutputFile(
Compaction Support for Range Deletion Summary: This diff introduces RangeDelAggregator, which takes ownership of iterators provided to it via AddTombstones(). The tombstones are organized in a two-level map (snapshot stripe -> begin key -> tombstone). Tombstone creation avoids data copy by holding Slices returned by the iterator, which remain valid thanks to pinning. For compaction, we create a hierarchical range tombstone iterator with structure matching the iterator over compaction input data. An aggregator based on that iterator is used by CompactionIterator to determine which keys are covered by range tombstones. In case of merge operand, the same aggregator is used by MergeHelper. Upon finishing each file in the compaction, relevant range tombstones are added to the output file's range tombstone metablock and file boundaries are updated accordingly. To check whether a key is covered by range tombstone, RangeDelAggregator::ShouldDelete() considers tombstones in the key's snapshot stripe. When this function is used outside of compaction, it also checks newer stripes, which can contain covering tombstones. Currently the intra-stripe check involves a linear scan; however, in the future we plan to collapse ranges within a stripe such that binary search can be used. RangeDelAggregator::AddToBuilder() adds all range tombstones in the table's key-range to a new table's range tombstone meta-block. Since range tombstones may fall in the gap between files, we may need to extend some files' key-ranges. The strategy is (1) first file extends as far left as possible and other files do not extend left, (2) all files extend right until either the start of the next file or the end of the last range tombstone in the gap, whichever comes first. One other notable change is adding release/move semantics to ScopedArenaIterator such that it can be used to transfer ownership of an arena-allocated iterator, similar to how unique_ptr is used for malloc'd data. Depends on D61473 Test Plan: compaction_iterator_test, mock_table, end-to-end tests in D63927 Reviewers: sdong, IslamAbdelRahman, wanning, yhchiang, lightmark Reviewed By: lightmark Subscribers: andrewkr, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D62205
8 years ago
const Status& input_status, SubcompactionState* sub_compact,
CompactionOutputs& outputs, const Slice& next_table_min_key) {
Allow GetThreadList() to report operation stage. Summary: Allow GetThreadList() to report operation stage. Test Plan: ./thread_list_test ./db_bench --benchmarks=fillrandom --num=100000 --threads=40 \ --max_background_compactions=10 --max_background_flushes=3 \ --thread_status_per_interval=1000 --key_size=16 --value_size=1000 \ --num_column_families=10 export ROCKSDB_TESTS=ThreadStatus ./db_test Sample output ThreadID ThreadType cfName Operation OP_StartTime ElapsedTime Stage State 140116265861184 Low Pri 140116270055488 Low Pri 140116274249792 High Pri column_family_name_000005 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116400078912 Low Pri column_family_name_000004 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116358135872 Low Pri column_family_name_000006 Compaction 2015/03/10-14:58:10 1 us CompactionJob::FinishCompactionOutputFile 140116341358656 Low Pri 140116295221312 High Pri default Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116324581440 Low Pri column_family_name_000009 Compaction 2015/03/10-14:58:11 0 us CompactionJob::ProcessKeyValueCompaction 140116278444096 Low Pri 140116299415616 Low Pri column_family_name_000008 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116291027008 High Pri column_family_name_000001 Flush 2015/03/10-14:58:11 0 us FlushJob::WriteLevel0Table 140116286832704 Low Pri column_family_name_000002 Compaction 2015/03/10-14:58:11 0 us CompactionJob::FinishCompactionOutputFile 140116282638400 Low Pri Reviewers: rven, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D34683
10 years ago
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_SYNC_FILE);
assert(sub_compact != nullptr);
assert(outputs.HasBuilder());
FileMetaData* meta = outputs.GetMetaData();
uint64_t output_number = meta->fd.GetNumber();
assert(output_number != 0);
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
std::string file_checksum = kUnknownFileChecksum;
std::string file_checksum_func_name = kUnknownFileChecksumFuncName;
// Check for iterator errors
Status s = input_status;
// Add range tombstones
auto earliest_snapshot = kMaxSequenceNumber;
if (existing_snapshots_.size() > 0) {
earliest_snapshot = existing_snapshots_[0];
}
if (s.ok()) {
CompactionIterationStats range_del_out_stats;
// if the compaction supports per_key_placement, only output range dels to
// the penultimate level.
// Note: Use `bottommost_level_ = true` for both bottommost and
// output_to_penultimate_level compaction here, as it's only used to decide
// if range dels could be dropped.
if (outputs.HasRangeDel()) {
s = outputs.AddRangeDels(
sub_compact->start.has_value() ? &(sub_compact->start.value())
: nullptr,
sub_compact->end.has_value() ? &(sub_compact->end.value()) : nullptr,
range_del_out_stats, bottommost_level_, cfd->internal_comparator(),
User-defined timestamp support for `DeleteRange()` (#10661) Summary: Add user-defined timestamp support for range deletion. The new API is `DeleteRange(opt, cf, begin_key, end_key, ts)`. Most of the change is to update the comparator to compare without timestamp. Other than that, major changes are - internal range tombstone data structures (`FragmentedRangeTombstoneList`, `RangeTombstone`, etc.) to store timestamps. - Garbage collection of range tombstones and range tombstone covered keys during compaction. - Get()/MultiGet() to return the timestamp of a range tombstone when needed. - Get/Iterator with range tombstones bounded by readoptions.timestamp. - timestamp crash test now issues DeleteRange by default. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10661 Test Plan: - Added unit test: `make check` - Stress test: `python3 tools/db_crashtest.py --enable_ts whitebox --readpercent=57 --prefixpercent=4 --writepercent=25 -delpercent=5 --iterpercent=5 --delrangepercent=4` - Ran `db_bench` to measure regression when timestamp is not enabled. The tests are for write (with some range deletion) and iterate with DB fitting in memory: `./db_bench--benchmarks=fillrandom,seekrandom --writes_per_range_tombstone=200 --max_write_buffer_number=100 --min_write_buffer_number_to_merge=100 --writes=500000 --reads=500000 --seek_nexts=10 --disable_auto_compactions -disable_wal=true --max_num_range_tombstones=1000`. Did not see consistent regression in no timestamp case. | micros/op | fillrandom | seekrandom | | --- | --- | --- | |main| 2.58 |10.96| |PR 10661| 2.68 |10.63| Reviewed By: riversand963 Differential Revision: D39441192 Pulled By: cbi42 fbshipit-source-id: f05aca3c41605caf110daf0ff405919f300ddec2
2 years ago
earliest_snapshot, next_table_min_key, full_history_ts_low_);
}
RecordDroppedKeys(range_del_out_stats, &sub_compact->compaction_job_stats);
TEST_SYNC_POINT("CompactionJob::FinishCompactionOutputFile1");
}
const uint64_t current_entries = outputs.NumEntries();
s = outputs.Finish(s, seqno_time_mapping_);
if (s.ok()) {
Try to start TTL earlier with kMinOverlappingRatio is used (#8749) Summary: Right now, when options.ttl is set, compactions are triggered around the time when TTL is reached. This might cause extra compactions which are often bursty. This commit tries to mitigate it by picking those files earlier in normal compaction picking process. This is only implemented using kMinOverlappingRatio with Leveled compaction as it is the default value and it is more complicated to change other styles. When a file is aged more than ttl/2, RocksDB starts to boost the compaction priority of files in normal compaction picking process, and hope by the time TTL is reached, very few extra compaction is needed. In order for this to work, another change is made: during a compaction, if an output level file is older than ttl/2, cut output files based on original boundary (if it is not in the last level). This is to make sure that after an old file is moved to the next level, and new data is merged from the upper level, the new data falling into this range isn't reset with old timestamp. Without this change, in many cases, most files from one level will keep having old timestamp, even if they have newer data and we stuck in it. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8749 Test Plan: Add a unit test to test the boosting logic. Will add a unit test to test it end-to-end. Reviewed By: jay-zhuang Differential Revision: D30735261 fbshipit-source-id: 503c2d89250b22911eb99e72b379be154de3428e
3 years ago
// With accurate smallest and largest key, we can get a slightly more
// accurate oldest ancester time.
// This makes oldest ancester time in manifest more accurate than in
// table properties. Not sure how to resolve it.
if (meta->smallest.size() > 0 && meta->largest.size() > 0) {
uint64_t refined_oldest_ancester_time;
Slice new_smallest = meta->smallest.user_key();
Slice new_largest = meta->largest.user_key();
if (!new_largest.empty() && !new_smallest.empty()) {
refined_oldest_ancester_time =
sub_compact->compaction->MinInputFileOldestAncesterTime(
&(meta->smallest), &(meta->largest));
if (refined_oldest_ancester_time !=
std::numeric_limits<uint64_t>::max()) {
Try to start TTL earlier with kMinOverlappingRatio is used (#8749) Summary: Right now, when options.ttl is set, compactions are triggered around the time when TTL is reached. This might cause extra compactions which are often bursty. This commit tries to mitigate it by picking those files earlier in normal compaction picking process. This is only implemented using kMinOverlappingRatio with Leveled compaction as it is the default value and it is more complicated to change other styles. When a file is aged more than ttl/2, RocksDB starts to boost the compaction priority of files in normal compaction picking process, and hope by the time TTL is reached, very few extra compaction is needed. In order for this to work, another change is made: during a compaction, if an output level file is older than ttl/2, cut output files based on original boundary (if it is not in the last level). This is to make sure that after an old file is moved to the next level, and new data is merged from the upper level, the new data falling into this range isn't reset with old timestamp. Without this change, in many cases, most files from one level will keep having old timestamp, even if they have newer data and we stuck in it. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8749 Test Plan: Add a unit test to test the boosting logic. Will add a unit test to test it end-to-end. Reviewed By: jay-zhuang Differential Revision: D30735261 fbshipit-source-id: 503c2d89250b22911eb99e72b379be154de3428e
3 years ago
meta->oldest_ancester_time = refined_oldest_ancester_time;
}
}
}
}
// Finish and check for file errors
IOStatus io_s = outputs.WriterSyncClose(s, db_options_.clock, stats_,
db_options_.use_fsync);
if (s.ok() && io_s.ok()) {
file_checksum = meta->file_checksum;
file_checksum_func_name = meta->file_checksum_func_name;
}
if (s.ok()) {
s = io_s;
}
if (sub_compact->io_status.ok()) {
sub_compact->io_status = io_s;
// Since this error is really a copy of the
// "normal" status, it does not also need to be checked
sub_compact->io_status.PermitUncheckedError();
}
TableProperties tp;
if (s.ok()) {
tp = outputs.GetTableProperties();
}
if (s.ok() && current_entries == 0 && tp.num_range_deletions == 0) {
// If there is nothing to output, no necessary to generate a sst file.
// This happens when the output level is bottom level, at the same time
// the sub_compact output nothing.
std::string fname =
TableFileName(sub_compact->compaction->immutable_options()->cf_paths,
meta->fd.GetNumber(), meta->fd.GetPathId());
// TODO(AR) it is not clear if there are any larger implications if
// DeleteFile fails here
Status ds = env_->DeleteFile(fname);
if (!ds.ok()) {
ROCKS_LOG_WARN(
db_options_.info_log,
"[%s] [JOB %d] Unable to remove SST file for table #%" PRIu64
" at bottom level%s",
cfd->GetName().c_str(), job_id_, output_number,
meta->marked_for_compaction ? " (need compaction)" : "");
}
// Also need to remove the file from outputs, or it will be added to the
// VersionEdit.
outputs.RemoveLastOutput();
meta = nullptr;
}
if (s.ok() && (current_entries > 0 || tp.num_range_deletions > 0)) {
// Output to event logger and fire events.
outputs.UpdateTableProperties();
ROCKS_LOG_INFO(db_options_.info_log,
"[%s] [JOB %d] Generated table #%" PRIu64 ": %" PRIu64
" keys, %" PRIu64 " bytes%s, temperature: %s",
cfd->GetName().c_str(), job_id_, output_number,
current_entries, meta->fd.file_size,
meta->marked_for_compaction ? " (need compaction)" : "",
temperature_to_string[meta->temperature].c_str());
}
std::string fname;
FileDescriptor output_fd;
uint64_t oldest_blob_file_number = kInvalidBlobFileNumber;
Status status_for_listener = s;
if (meta != nullptr) {
fname = GetTableFileName(meta->fd.GetNumber());
output_fd = meta->fd;
oldest_blob_file_number = meta->oldest_blob_file_number;
} else {
fname = "(nil)";
if (s.ok()) {
status_for_listener = Status::Aborted("Empty SST file not kept");
}
}
EventHelpers::LogAndNotifyTableFileCreationFinished(
event_logger_, cfd->ioptions()->listeners, dbname_, cfd->GetName(), fname,
job_id_, output_fd, oldest_blob_file_number, tp,
TableFileCreationReason::kCompaction, status_for_listener, file_checksum,
file_checksum_func_name);
// Report new file to SstFileManagerImpl
auto sfm =
static_cast<SstFileManagerImpl*>(db_options_.sst_file_manager.get());
if (sfm && meta != nullptr && meta->fd.GetPathId() == 0) {
Status add_s = sfm->OnAddFile(fname);
if (!add_s.ok() && s.ok()) {
s = add_s;
}
if (sfm->IsMaxAllowedSpaceReached()) {
// TODO(ajkr): should we return OK() if max space was reached by the final
// compaction output file (similarly to how flush works when full)?
s = Status::SpaceLimit("Max allowed space was reached");
TEST_SYNC_POINT(
"CompactionJob::FinishCompactionOutputFile:MaxAllowedSpaceReached");
InstrumentedMutexLock l(db_mutex_);
db_error_handler_->SetBGError(s, BackgroundErrorReason::kCompaction);
}
}
outputs.ResetBuilder();
return s;
}
Status CompactionJob::InstallCompactionResults(
const MutableCFOptions& mutable_cf_options) {
assert(compact_);
db_mutex_->AssertHeld();
auto* compaction = compact_->compaction;
assert(compaction);
{
Compaction::InputLevelSummaryBuffer inputs_summary;
if (compaction_stats_.has_penultimate_level_output) {
ROCKS_LOG_BUFFER(
log_buffer_,
"[%s] [JOB %d] Compacted %s => output_to_penultimate_level: %" PRIu64
" bytes + last: %" PRIu64 " bytes. Total: %" PRIu64 " bytes",
compaction->column_family_data()->GetName().c_str(), job_id_,
compaction->InputLevelSummary(&inputs_summary),
compaction_stats_.penultimate_level_stats.bytes_written,
compaction_stats_.stats.bytes_written,
compaction_stats_.TotalBytesWritten());
} else {
ROCKS_LOG_BUFFER(log_buffer_,
"[%s] [JOB %d] Compacted %s => %" PRIu64 " bytes",
compaction->column_family_data()->GetName().c_str(),
job_id_, compaction->InputLevelSummary(&inputs_summary),
compaction_stats_.TotalBytesWritten());
}
}
VersionEdit* const edit = compaction->edit();
assert(edit);
// Add compaction inputs
compaction->AddInputDeletions(edit);
std::unordered_map<uint64_t, BlobGarbageMeter::BlobStats> blob_total_garbage;
for (const auto& sub_compact : compact_->sub_compact_states) {
sub_compact.AddOutputsEdit(edit);
for (const auto& blob : sub_compact.Current().GetBlobFileAdditions()) {
edit->AddBlobFile(blob);
}
if (sub_compact.Current().GetBlobGarbageMeter()) {
const auto& flows = sub_compact.Current().GetBlobGarbageMeter()->flows();
for (const auto& pair : flows) {
const uint64_t blob_file_number = pair.first;
const BlobGarbageMeter::BlobInOutFlow& flow = pair.second;
assert(flow.IsValid());
if (flow.HasGarbage()) {
blob_total_garbage[blob_file_number].Add(flow.GetGarbageCount(),
flow.GetGarbageBytes());
}
}
}
}
for (const auto& pair : blob_total_garbage) {
const uint64_t blob_file_number = pair.first;
const BlobGarbageMeter::BlobStats& stats = pair.second;
edit->AddBlobFileGarbage(blob_file_number, stats.GetCount(),
stats.GetBytes());
}
if ((compaction->compaction_reason() ==
CompactionReason::kLevelMaxLevelSize ||
compaction->compaction_reason() == CompactionReason::kRoundRobinTtl) &&
Add basic kRoundRobin compaction policy (#10107) Summary: Add `kRoundRobin` as a compaction priority. The implementation is as follows. - Define a cursor as the smallest Internal key in the successor of the selected file. Add `vector<InternalKey> compact_cursor_` into `VersionStorageInfo` where each element (`InternalKey`) in `compact_cursor_` represents a cursor. In round-robin compaction policy, we just need to select the first file (assuming files are sorted) and also has the smallest InternalKey larger than/equal to the cursor. After a file is chosen, we create a new `Fsize` vector which puts the selected file is placed at the first position in `temp`, the next cursor is then updated as the smallest InternalKey in successor of the selected file (the above logic is implemented in `SortFileByRoundRobin`). - After a compaction succeeds, typically `InstallCompactionResults()`, we choose the next cursor for the input level and save it to `edit`. When calling `LogAndApply`, we save the next cursor with its level into some local variable and finally apply the change to `vstorage` in `SaveTo` function. - Cursors are persist pair by pair (<level, InternalKey>) in `EncodeTo` so that they can be reconstructed when reopening. An empty cursor will not be encoded to MANIFEST Pull Request resolved: https://github.com/facebook/rocksdb/pull/10107 Test Plan: add unit test (`CompactionPriRoundRobin`) in `compaction_picker_test`, add `kRoundRobin` priority in `CompactionPriTest` from `db_compaction_test`, and add `PersistRoundRobinCompactCursor` in `db_compaction_test` Reviewed By: ajkr Differential Revision: D37316037 Pulled By: littlepig2013 fbshipit-source-id: 9f481748190ace416079139044e00df2968fb1ee
3 years ago
compaction->immutable_options()->compaction_pri == kRoundRobin) {
int start_level = compaction->start_level();
if (start_level > 0) {
auto vstorage = compaction->input_version()->storage_info();
edit->AddCompactCursor(start_level,
Support subcmpct using reserved resources for round-robin priority (#10341) Summary: Earlier implementation of round-robin priority can only pick one file at a time and disallows parallel compactions within the same level. In this PR, round-robin compaction policy will expand towards more input files with respecting some additional constraints, which are summarized as follows: * Constraint 1: We can only pick consecutive files - Constraint 1a: When a file is being compacted (or some input files are being compacted after expanding), we cannot choose it and have to stop choosing more files - Constraint 1b: When we reach the last file (with the largest keys), we cannot choose more files (the next file will be the first one with small keys) * Constraint 2: We should ensure the total compaction bytes (including the overlapped files from the next level) is no more than `mutable_cf_options_.max_compaction_bytes` * Constraint 3: We try our best to pick as many files as possible so that the post-compaction level size can be just less than `MaxBytesForLevel(start_level_)` * Constraint 4: If trivial move is allowed, we reuse the logic of `TryNonL0TrivialMove()` instead of expanding files with Constraint 3 More details can be found in `LevelCompactionBuilder::SetupOtherFilesWithRoundRobinExpansion()`. The above optimization accelerates the process of moving the compaction cursor, in which the write-amp can be further reduced. While a large compaction may lead to high write stall, we break this large compaction into several subcompactions **regardless of** the `max_subcompactions` limit. The number of subcompactions for round-robin compaction priority is determined through the following steps: * Step 1: Initialized against `max_output_file_limit`, the number of input files in the start level, and also the range size limit `ranges.size()` * Step 2: Call `AcquireSubcompactionResources()`when max subcompactions is not sufficient, but we may or may not obtain desired resources, additional number of resources is stored in `extra_num_subcompaction_threads_reserved_`). Subcompaction limit is changed and update `num_planned_subcompactions` with `GetSubcompactionLimit()` * Step 3: Call `ShrinkSubcompactionResources()` to ensure extra resources can be released (extra resources may exist for round-robin compaction when the number of actual number of subcompactions is less than the number of planned subcompactions) More details can be found in `CompactionJob::AcquireSubcompactionResources()`,`CompactionJob::ShrinkSubcompactionResources()`, and `CompactionJob::ReleaseSubcompactionResources()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/10341 Test Plan: Add `CompactionPriMultipleFilesRoundRobin[1-3]` unit test in `compaction_picker_test.cc` and `RoundRobinSubcompactionsAgainstResources.SubcompactionsUsingResources/[0-4]`, `RoundRobinSubcompactionsAgainstPressureToken.PressureTokenTest/[0-1]` in `db_compaction_test.cc` Reviewed By: ajkr, hx235 Differential Revision: D37792644 Pulled By: littlepig2013 fbshipit-source-id: 7fecb7c4ffd97b34bbf6e3b760b2c35a772a0657
2 years ago
vstorage->GetNextCompactCursor(
start_level, compaction->num_input_files(0)));
Add basic kRoundRobin compaction policy (#10107) Summary: Add `kRoundRobin` as a compaction priority. The implementation is as follows. - Define a cursor as the smallest Internal key in the successor of the selected file. Add `vector<InternalKey> compact_cursor_` into `VersionStorageInfo` where each element (`InternalKey`) in `compact_cursor_` represents a cursor. In round-robin compaction policy, we just need to select the first file (assuming files are sorted) and also has the smallest InternalKey larger than/equal to the cursor. After a file is chosen, we create a new `Fsize` vector which puts the selected file is placed at the first position in `temp`, the next cursor is then updated as the smallest InternalKey in successor of the selected file (the above logic is implemented in `SortFileByRoundRobin`). - After a compaction succeeds, typically `InstallCompactionResults()`, we choose the next cursor for the input level and save it to `edit`. When calling `LogAndApply`, we save the next cursor with its level into some local variable and finally apply the change to `vstorage` in `SaveTo` function. - Cursors are persist pair by pair (<level, InternalKey>) in `EncodeTo` so that they can be reconstructed when reopening. An empty cursor will not be encoded to MANIFEST Pull Request resolved: https://github.com/facebook/rocksdb/pull/10107 Test Plan: add unit test (`CompactionPriRoundRobin`) in `compaction_picker_test`, add `kRoundRobin` priority in `CompactionPriTest` from `db_compaction_test`, and add `PersistRoundRobinCompactCursor` in `db_compaction_test` Reviewed By: ajkr Differential Revision: D37316037 Pulled By: littlepig2013 fbshipit-source-id: 9f481748190ace416079139044e00df2968fb1ee
3 years ago
}
}
return versions_->LogAndApply(compaction->column_family_data(),
mutable_cf_options, edit, db_mutex_,
db_directory_);
}
void CompactionJob::RecordCompactionIOStats() {
RecordTick(stats_, COMPACT_READ_BYTES, IOSTATS(bytes_read));
RecordTick(stats_, COMPACT_WRITE_BYTES, IOSTATS(bytes_written));
CompactionReason compaction_reason =
compact_->compaction->compaction_reason();
if (compaction_reason == CompactionReason::kFilesMarkedForCompaction) {
RecordTick(stats_, COMPACT_READ_BYTES_MARKED, IOSTATS(bytes_read));
RecordTick(stats_, COMPACT_WRITE_BYTES_MARKED, IOSTATS(bytes_written));
} else if (compaction_reason == CompactionReason::kPeriodicCompaction) {
RecordTick(stats_, COMPACT_READ_BYTES_PERIODIC, IOSTATS(bytes_read));
RecordTick(stats_, COMPACT_WRITE_BYTES_PERIODIC, IOSTATS(bytes_written));
} else if (compaction_reason == CompactionReason::kTtl) {
RecordTick(stats_, COMPACT_READ_BYTES_TTL, IOSTATS(bytes_read));
RecordTick(stats_, COMPACT_WRITE_BYTES_TTL, IOSTATS(bytes_written));
}
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ThreadStatusUtil::IncreaseThreadOperationProperty(
ThreadStatus::COMPACTION_BYTES_READ, IOSTATS(bytes_read));
IOSTATS_RESET(bytes_read);
Allow GetThreadList() to report basic compaction operation properties. Summary: Now we're able to show more details about a compaction in GetThreadList() :) This patch allows GetThreadList() to report basic compaction operation properties. Basic compaction properties include: 1. job id 2. compaction input / output level 3. compaction property flags (is_manual, is_deletion, .. etc) 4. total input bytes 5. the number of bytes has been read currently. 6. the number of bytes has been written currently. Flush operation properties will be done in a seperate diff. Test Plan: /db_bench --threads=30 --num=1000000 --benchmarks=fillrandom --thread_status_per_interval=1 Sample output of tracking same job: ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 31.357 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 59.440 ms CompactionJob::FinishCompactionOutputFile BaseInputLevel 1 | BytesRead 2264663 | BytesWritten 1934241 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | ThreadID ThreadType cfName Operation ElapsedTime Stage State OperationProperties 140664171987072 Low Pri default Compaction 226.375 ms CompactionJob::Install BaseInputLevel 1 | BytesRead 3958013 | BytesWritten 3621940 | IsDeletion 0 | IsManual 0 | IsTrivialMove 0 | JobID 277 | OutputLevel 2 | TotalInputBytes 3964158 | Reviewers: sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D37653
10 years ago
ThreadStatusUtil::IncreaseThreadOperationProperty(
ThreadStatus::COMPACTION_BYTES_WRITTEN, IOSTATS(bytes_written));
IOSTATS_RESET(bytes_written);
}
Status CompactionJob::OpenCompactionOutputFile(SubcompactionState* sub_compact,
CompactionOutputs& outputs) {
assert(sub_compact != nullptr);
// no need to lock because VersionSet::next_file_number_ is atomic
uint64_t file_number = versions_->NewFileNumber();
std::string fname = GetTableFileName(file_number);
// Fire events.
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
EventHelpers::NotifyTableFileCreationStarted(
cfd->ioptions()->listeners, dbname_, cfd->GetName(), fname, job_id_,
TableFileCreationReason::kCompaction);
// Make the output file
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
std::unique_ptr<FSWritableFile> writable_file;
#ifndef NDEBUG
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
bool syncpoint_arg = file_options_.use_direct_writes;
TEST_SYNC_POINT_CALLBACK("CompactionJob::OpenCompactionOutputFile",
&syncpoint_arg);
#endif
// Pass temperature of the last level files to FileSystem.
FileOptions fo_copy = file_options_;
Temperature temperature = sub_compact->compaction->output_temperature();
// only set for the last level compaction and also it's not output to
// penultimate level (when preclude_last_level feature is enabled)
if (temperature == Temperature::kUnknown &&
sub_compact->compaction->is_last_level() &&
!sub_compact->IsCurrentPenultimateLevel()) {
temperature =
sub_compact->compaction->mutable_cf_options()->last_level_temperature;
}
fo_copy.temperature = temperature;
Status s;
IOStatus io_s = NewWritableFile(fs_.get(), fname, &writable_file, fo_copy);
s = io_s;
if (sub_compact->io_status.ok()) {
sub_compact->io_status = io_s;
// Since this error is really a copy of the io_s that is checked below as s,
// it does not also need to be checked.
sub_compact->io_status.PermitUncheckedError();
}
if (!s.ok()) {
ROCKS_LOG_ERROR(
db_options_.info_log,
"[%s] [JOB %d] OpenCompactionOutputFiles for table #%" PRIu64
" fails at NewWritableFile with status %s",
sub_compact->compaction->column_family_data()->GetName().c_str(),
job_id_, file_number, s.ToString().c_str());
LogFlush(db_options_.info_log);
EventHelpers::LogAndNotifyTableFileCreationFinished(
event_logger_, cfd->ioptions()->listeners, dbname_, cfd->GetName(),
fname, job_id_, FileDescriptor(), kInvalidBlobFileNumber,
TableProperties(), TableFileCreationReason::kCompaction, s,
kUnknownFileChecksum, kUnknownFileChecksumFuncName);
return s;
}
// Try to figure out the output file's oldest ancester time.
int64_t temp_current_time = 0;
auto get_time_status = db_options_.clock->GetCurrentTime(&temp_current_time);
// Safe to proceed even if GetCurrentTime fails. So, log and proceed.
if (!get_time_status.ok()) {
ROCKS_LOG_WARN(db_options_.info_log,
"Failed to get current time. Status: %s",
get_time_status.ToString().c_str());
}
uint64_t current_time = static_cast<uint64_t>(temp_current_time);
Try to start TTL earlier with kMinOverlappingRatio is used (#8749) Summary: Right now, when options.ttl is set, compactions are triggered around the time when TTL is reached. This might cause extra compactions which are often bursty. This commit tries to mitigate it by picking those files earlier in normal compaction picking process. This is only implemented using kMinOverlappingRatio with Leveled compaction as it is the default value and it is more complicated to change other styles. When a file is aged more than ttl/2, RocksDB starts to boost the compaction priority of files in normal compaction picking process, and hope by the time TTL is reached, very few extra compaction is needed. In order for this to work, another change is made: during a compaction, if an output level file is older than ttl/2, cut output files based on original boundary (if it is not in the last level). This is to make sure that after an old file is moved to the next level, and new data is merged from the upper level, the new data falling into this range isn't reset with old timestamp. Without this change, in many cases, most files from one level will keep having old timestamp, even if they have newer data and we stuck in it. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8749 Test Plan: Add a unit test to test the boosting logic. Will add a unit test to test it end-to-end. Reviewed By: jay-zhuang Differential Revision: D30735261 fbshipit-source-id: 503c2d89250b22911eb99e72b379be154de3428e
3 years ago
InternalKey tmp_start, tmp_end;
if (sub_compact->start.has_value()) {
tmp_start.SetMinPossibleForUserKey(sub_compact->start.value());
Try to start TTL earlier with kMinOverlappingRatio is used (#8749) Summary: Right now, when options.ttl is set, compactions are triggered around the time when TTL is reached. This might cause extra compactions which are often bursty. This commit tries to mitigate it by picking those files earlier in normal compaction picking process. This is only implemented using kMinOverlappingRatio with Leveled compaction as it is the default value and it is more complicated to change other styles. When a file is aged more than ttl/2, RocksDB starts to boost the compaction priority of files in normal compaction picking process, and hope by the time TTL is reached, very few extra compaction is needed. In order for this to work, another change is made: during a compaction, if an output level file is older than ttl/2, cut output files based on original boundary (if it is not in the last level). This is to make sure that after an old file is moved to the next level, and new data is merged from the upper level, the new data falling into this range isn't reset with old timestamp. Without this change, in many cases, most files from one level will keep having old timestamp, even if they have newer data and we stuck in it. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8749 Test Plan: Add a unit test to test the boosting logic. Will add a unit test to test it end-to-end. Reviewed By: jay-zhuang Differential Revision: D30735261 fbshipit-source-id: 503c2d89250b22911eb99e72b379be154de3428e
3 years ago
}
if (sub_compact->end.has_value()) {
tmp_end.SetMinPossibleForUserKey(sub_compact->end.value());
Try to start TTL earlier with kMinOverlappingRatio is used (#8749) Summary: Right now, when options.ttl is set, compactions are triggered around the time when TTL is reached. This might cause extra compactions which are often bursty. This commit tries to mitigate it by picking those files earlier in normal compaction picking process. This is only implemented using kMinOverlappingRatio with Leveled compaction as it is the default value and it is more complicated to change other styles. When a file is aged more than ttl/2, RocksDB starts to boost the compaction priority of files in normal compaction picking process, and hope by the time TTL is reached, very few extra compaction is needed. In order for this to work, another change is made: during a compaction, if an output level file is older than ttl/2, cut output files based on original boundary (if it is not in the last level). This is to make sure that after an old file is moved to the next level, and new data is merged from the upper level, the new data falling into this range isn't reset with old timestamp. Without this change, in many cases, most files from one level will keep having old timestamp, even if they have newer data and we stuck in it. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8749 Test Plan: Add a unit test to test the boosting logic. Will add a unit test to test it end-to-end. Reviewed By: jay-zhuang Differential Revision: D30735261 fbshipit-source-id: 503c2d89250b22911eb99e72b379be154de3428e
3 years ago
}
uint64_t oldest_ancester_time =
Try to start TTL earlier with kMinOverlappingRatio is used (#8749) Summary: Right now, when options.ttl is set, compactions are triggered around the time when TTL is reached. This might cause extra compactions which are often bursty. This commit tries to mitigate it by picking those files earlier in normal compaction picking process. This is only implemented using kMinOverlappingRatio with Leveled compaction as it is the default value and it is more complicated to change other styles. When a file is aged more than ttl/2, RocksDB starts to boost the compaction priority of files in normal compaction picking process, and hope by the time TTL is reached, very few extra compaction is needed. In order for this to work, another change is made: during a compaction, if an output level file is older than ttl/2, cut output files based on original boundary (if it is not in the last level). This is to make sure that after an old file is moved to the next level, and new data is merged from the upper level, the new data falling into this range isn't reset with old timestamp. Without this change, in many cases, most files from one level will keep having old timestamp, even if they have newer data and we stuck in it. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8749 Test Plan: Add a unit test to test the boosting logic. Will add a unit test to test it end-to-end. Reviewed By: jay-zhuang Differential Revision: D30735261 fbshipit-source-id: 503c2d89250b22911eb99e72b379be154de3428e
3 years ago
sub_compact->compaction->MinInputFileOldestAncesterTime(
sub_compact->start.has_value() ? &tmp_start : nullptr,
sub_compact->end.has_value() ? &tmp_end : nullptr);
if (oldest_ancester_time == std::numeric_limits<uint64_t>::max()) {
oldest_ancester_time = current_time;
}
// Initialize a SubcompactionState::Output and add it to sub_compact->outputs
Sort L0 files by newly introduced epoch_num (#10922) Summary: **Context:** Sorting L0 files by `largest_seqno` has at least two inconvenience: - File ingestion and compaction involving ingested files can create files of overlapping seqno range with the existing files. `force_consistency_check=true` will catch such overlap seqno range even those harmless overlap. - For example, consider the following sequence of events ("key@n" indicates key at seqno "n") - insert k1@1 to memtable m1 - ingest file s1 with k2@2, ingest file s2 with k3@3 - insert k4@4 to m1 - compact files s1, s2 and result in new file s3 of seqno range [2, 3] - flush m1 and result in new file s4 of seqno range [1, 4]. And `force_consistency_check=true` will think s4 and s3 has file reordering corruption that might cause retuning an old value of k1 - However such caught corruption is a false positive since s1, s2 will not have overlapped keys with k1 or whatever inserted into m1 before ingest file s1 by the requirement of file ingestion (otherwise the m1 will be flushed first before any of the file ingestion completes). Therefore there in fact isn't any file reordering corruption. - Single delete can decrease a file's largest seqno and ordering by `largest_seqno` can introduce a wrong ordering hence file reordering corruption - For example, consider the following sequence of events ("key@n" indicates key at seqno "n", Credit to ajkr for this example) - an existing SST s1 contains only k1@1 - insert k1@2 to memtable m1 - ingest file s2 with k3@3, ingest file s3 with k4@4 - insert single delete k5@5 in m1 - flush m1 and result in new file s4 of seqno range [2, 5] - compact s1, s2, s3 and result in new file s5 of seqno range [1, 4] - compact s4 and result in new file s6 of seqno range [2] due to single delete - By the last step, we have file ordering by largest seqno (">" means "newer") : s5 > s6 while s6 contains a newer version of the k1's value (i.e, k1@2) than s5, which is a real reordering corruption. While this can be caught by `force_consistency_check=true`, there isn't a good way to prevent this from happening if ordering by `largest_seqno` Therefore, we are redesigning the sorting criteria of L0 files and avoid above inconvenience. Credit to ajkr , we now introduce `epoch_num` which describes the order of a file being flushed or ingested/imported (compaction output file will has the minimum `epoch_num` among input files'). This will avoid the above inconvenience in the following ways: - In the first case above, there will no longer be overlap seqno range check in `force_consistency_check=true` but `epoch_number` ordering check. This will result in file ordering s1 < s2 < s4 (pre-compaction) and s3 < s4 (post-compaction) which won't trigger false positive corruption. See test class `DBCompactionTestL0FilesMisorderCorruption*` for more. - In the second case above, this will result in file ordering s1 < s2 < s3 < s4 (pre-compacting s1, s2, s3), s5 < s4 (post-compacting s1, s2, s3), s5 < s6 (post-compacting s4), which are correct file ordering without causing any corruption. **Summary:** - Introduce `epoch_number` stored per `ColumnFamilyData` and sort CF's L0 files by their assigned `epoch_number` instead of `largest_seqno`. - `epoch_number` is increased and assigned upon `VersionEdit::AddFile()` for flush (or similarly for WriteLevel0TableForRecovery) and file ingestion (except for allow_behind_true, which will always get assigned as the `kReservedEpochNumberForFileIngestedBehind`) - Compaction output file is assigned with the minimum `epoch_number` among input files' - Refit level: reuse refitted file's epoch_number - Other paths needing `epoch_number` treatment: - Import column families: reuse file's epoch_number if exists. If not, assign one based on `NewestFirstBySeqNo` - Repair: reuse file's epoch_number if exists. If not, assign one based on `NewestFirstBySeqNo`. - Assigning new epoch_number to a file and adding this file to LSM tree should be atomic. This is guaranteed by us assigning epoch_number right upon `VersionEdit::AddFile()` where this version edit will be apply to LSM tree shape right after by holding the db mutex (e.g, flush, file ingestion, import column family) or by there is only 1 ongoing edit per CF (e.g, WriteLevel0TableForRecovery, Repair). - Assigning the minimum input epoch number to compaction output file won't misorder L0 files (even through later `Refit(target_level=0)`). It's due to for every key "k" in the input range, a legit compaction will cover a continuous epoch number range of that key. As long as we assign the key "k" the minimum input epoch number, it won't become newer or older than the versions of this key that aren't included in this compaction hence no misorder. - Persist `epoch_number` of each file in manifest and recover `epoch_number` on db recovery - Backward compatibility with old db without `epoch_number` support is guaranteed by assigning `epoch_number` to recovered files by `NewestFirstBySeqno` order. See `VersionStorageInfo::RecoverEpochNumbers()` for more - Forward compatibility with manifest is guaranteed by flexibility of `NewFileCustomTag` - Replace `force_consistent_check` on L0 with `epoch_number` and remove false positive check like case 1 with `largest_seqno` above - Due to backward compatibility issue, we might encounter files with missing epoch number at the beginning of db recovery. We will still use old L0 sorting mechanism (`NewestFirstBySeqno`) to check/sort them till we infer their epoch number. See usages of `EpochNumberRequirement`. - Remove fix https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and their outdated tests to file reordering corruption because such fix can be replaced by this PR. - Misc: - update existing tests with `epoch_number` so make check will pass - update https://github.com/facebook/rocksdb/pull/5958#issue-511150930 tests to verify corruption is fixed using `epoch_number` and cover universal/fifo compaction/CompactRange/CompactFile cases - assert db_mutex is held for a few places before calling ColumnFamilyData::NewEpochNumber() Pull Request resolved: https://github.com/facebook/rocksdb/pull/10922 Test Plan: - `make check` - New unit tests under `db/db_compaction_test.cc`, `db/db_test2.cc`, `db/version_builder_test.cc`, `db/repair_test.cc` - Updated tests (i.e, `DBCompactionTestL0FilesMisorderCorruption*`) under https://github.com/facebook/rocksdb/pull/5958#issue-511150930 - [Ongoing] Compatibility test: manually run https://github.com/ajkr/rocksdb/commit/36a5686ec012f35a4371e409aa85c404ca1c210d (with file ingestion off for running the `.orig` binary to prevent this bug affecting upgrade/downgrade formality checking) for 1 hour on `simple black/white box`, `cf_consistency/txn/enable_ts with whitebox + test_best_efforts_recovery with blackbox` - [Ongoing] normal db stress test - [Ongoing] db stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 Reviewed By: ajkr Differential Revision: D41063187 Pulled By: hx235 fbshipit-source-id: 826cb23455de7beaabe2d16c57682a82733a32a9
2 years ago
uint64_t epoch_number = sub_compact->compaction->MinInputFileEpochNumber();
{
FileMetaData meta;
meta.fd = FileDescriptor(file_number,
sub_compact->compaction->output_path_id(), 0);
meta.oldest_ancester_time = oldest_ancester_time;
meta.file_creation_time = current_time;
Sort L0 files by newly introduced epoch_num (#10922) Summary: **Context:** Sorting L0 files by `largest_seqno` has at least two inconvenience: - File ingestion and compaction involving ingested files can create files of overlapping seqno range with the existing files. `force_consistency_check=true` will catch such overlap seqno range even those harmless overlap. - For example, consider the following sequence of events ("key@n" indicates key at seqno "n") - insert k1@1 to memtable m1 - ingest file s1 with k2@2, ingest file s2 with k3@3 - insert k4@4 to m1 - compact files s1, s2 and result in new file s3 of seqno range [2, 3] - flush m1 and result in new file s4 of seqno range [1, 4]. And `force_consistency_check=true` will think s4 and s3 has file reordering corruption that might cause retuning an old value of k1 - However such caught corruption is a false positive since s1, s2 will not have overlapped keys with k1 or whatever inserted into m1 before ingest file s1 by the requirement of file ingestion (otherwise the m1 will be flushed first before any of the file ingestion completes). Therefore there in fact isn't any file reordering corruption. - Single delete can decrease a file's largest seqno and ordering by `largest_seqno` can introduce a wrong ordering hence file reordering corruption - For example, consider the following sequence of events ("key@n" indicates key at seqno "n", Credit to ajkr for this example) - an existing SST s1 contains only k1@1 - insert k1@2 to memtable m1 - ingest file s2 with k3@3, ingest file s3 with k4@4 - insert single delete k5@5 in m1 - flush m1 and result in new file s4 of seqno range [2, 5] - compact s1, s2, s3 and result in new file s5 of seqno range [1, 4] - compact s4 and result in new file s6 of seqno range [2] due to single delete - By the last step, we have file ordering by largest seqno (">" means "newer") : s5 > s6 while s6 contains a newer version of the k1's value (i.e, k1@2) than s5, which is a real reordering corruption. While this can be caught by `force_consistency_check=true`, there isn't a good way to prevent this from happening if ordering by `largest_seqno` Therefore, we are redesigning the sorting criteria of L0 files and avoid above inconvenience. Credit to ajkr , we now introduce `epoch_num` which describes the order of a file being flushed or ingested/imported (compaction output file will has the minimum `epoch_num` among input files'). This will avoid the above inconvenience in the following ways: - In the first case above, there will no longer be overlap seqno range check in `force_consistency_check=true` but `epoch_number` ordering check. This will result in file ordering s1 < s2 < s4 (pre-compaction) and s3 < s4 (post-compaction) which won't trigger false positive corruption. See test class `DBCompactionTestL0FilesMisorderCorruption*` for more. - In the second case above, this will result in file ordering s1 < s2 < s3 < s4 (pre-compacting s1, s2, s3), s5 < s4 (post-compacting s1, s2, s3), s5 < s6 (post-compacting s4), which are correct file ordering without causing any corruption. **Summary:** - Introduce `epoch_number` stored per `ColumnFamilyData` and sort CF's L0 files by their assigned `epoch_number` instead of `largest_seqno`. - `epoch_number` is increased and assigned upon `VersionEdit::AddFile()` for flush (or similarly for WriteLevel0TableForRecovery) and file ingestion (except for allow_behind_true, which will always get assigned as the `kReservedEpochNumberForFileIngestedBehind`) - Compaction output file is assigned with the minimum `epoch_number` among input files' - Refit level: reuse refitted file's epoch_number - Other paths needing `epoch_number` treatment: - Import column families: reuse file's epoch_number if exists. If not, assign one based on `NewestFirstBySeqNo` - Repair: reuse file's epoch_number if exists. If not, assign one based on `NewestFirstBySeqNo`. - Assigning new epoch_number to a file and adding this file to LSM tree should be atomic. This is guaranteed by us assigning epoch_number right upon `VersionEdit::AddFile()` where this version edit will be apply to LSM tree shape right after by holding the db mutex (e.g, flush, file ingestion, import column family) or by there is only 1 ongoing edit per CF (e.g, WriteLevel0TableForRecovery, Repair). - Assigning the minimum input epoch number to compaction output file won't misorder L0 files (even through later `Refit(target_level=0)`). It's due to for every key "k" in the input range, a legit compaction will cover a continuous epoch number range of that key. As long as we assign the key "k" the minimum input epoch number, it won't become newer or older than the versions of this key that aren't included in this compaction hence no misorder. - Persist `epoch_number` of each file in manifest and recover `epoch_number` on db recovery - Backward compatibility with old db without `epoch_number` support is guaranteed by assigning `epoch_number` to recovered files by `NewestFirstBySeqno` order. See `VersionStorageInfo::RecoverEpochNumbers()` for more - Forward compatibility with manifest is guaranteed by flexibility of `NewFileCustomTag` - Replace `force_consistent_check` on L0 with `epoch_number` and remove false positive check like case 1 with `largest_seqno` above - Due to backward compatibility issue, we might encounter files with missing epoch number at the beginning of db recovery. We will still use old L0 sorting mechanism (`NewestFirstBySeqno`) to check/sort them till we infer their epoch number. See usages of `EpochNumberRequirement`. - Remove fix https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and their outdated tests to file reordering corruption because such fix can be replaced by this PR. - Misc: - update existing tests with `epoch_number` so make check will pass - update https://github.com/facebook/rocksdb/pull/5958#issue-511150930 tests to verify corruption is fixed using `epoch_number` and cover universal/fifo compaction/CompactRange/CompactFile cases - assert db_mutex is held for a few places before calling ColumnFamilyData::NewEpochNumber() Pull Request resolved: https://github.com/facebook/rocksdb/pull/10922 Test Plan: - `make check` - New unit tests under `db/db_compaction_test.cc`, `db/db_test2.cc`, `db/version_builder_test.cc`, `db/repair_test.cc` - Updated tests (i.e, `DBCompactionTestL0FilesMisorderCorruption*`) under https://github.com/facebook/rocksdb/pull/5958#issue-511150930 - [Ongoing] Compatibility test: manually run https://github.com/ajkr/rocksdb/commit/36a5686ec012f35a4371e409aa85c404ca1c210d (with file ingestion off for running the `.orig` binary to prevent this bug affecting upgrade/downgrade formality checking) for 1 hour on `simple black/white box`, `cf_consistency/txn/enable_ts with whitebox + test_best_efforts_recovery with blackbox` - [Ongoing] normal db stress test - [Ongoing] db stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 Reviewed By: ajkr Differential Revision: D41063187 Pulled By: hx235 fbshipit-source-id: 826cb23455de7beaabe2d16c57682a82733a32a9
2 years ago
meta.epoch_number = epoch_number;
meta.temperature = temperature;
assert(!db_id_.empty());
assert(!db_session_id_.empty());
s = GetSstInternalUniqueId(db_id_, db_session_id_, meta.fd.GetNumber(),
&meta.unique_id);
if (!s.ok()) {
ROCKS_LOG_ERROR(db_options_.info_log,
"[%s] [JOB %d] file #%" PRIu64
" failed to generate unique id: %s.",
cfd->GetName().c_str(), job_id_, meta.fd.GetNumber(),
s.ToString().c_str());
return s;
}
outputs.AddOutput(std::move(meta), cfd->internal_comparator(),
sub_compact->compaction->mutable_cf_options()
->check_flush_compaction_key_order,
paranoid_file_checks_);
}
Set Write rate limiter priority dynamically and pass it to FS (#9988) Summary: ### Context: Background compactions and flush generate large reads and writes, and can be long running, especially for universal compaction. In some cases, this can impact foreground reads and writes by users. From the RocksDB perspective, there can be two kinds of rate limiters, the internal (native) one and the external one. - The internal (native) rate limiter is introduced in [the wiki](https://github.com/facebook/rocksdb/wiki/Rate-Limiter). Currently, only IO_LOW and IO_HIGH are used and they are set statically. - For the external rate limiter, in FSWritableFile functions, IOOptions is open for end users to set and get rate_limiter_priority for their own rate limiter. Currently, RocksDB doesn’t pass the rate_limiter_priority through IOOptions to the file system. ### Solution During the User Read, Flush write, Compaction read/write, the WriteController is used to determine whether DB writes are stalled or slowed down. The rate limiter priority (Env::IOPriority) can be determined accordingly. We decided to always pass the priority in IOOptions. What the file system does with it should be a contract between the user and the file system. We would like to set the rate limiter priority at file level, since the Flush/Compaction job level may be too coarse with multiple files and block IO level is too granular. **This PR is for the Write path.** The **Write:** dynamic priority for different state are listed as follows: | State | Normal | Delayed | Stalled | | ----- | ------ | ------- | ------- | | Flush | IO_HIGH | IO_USER | IO_USER | | Compaction | IO_LOW | IO_USER | IO_USER | Flush and Compaction writes share the same call path through BlockBaseTableWriter, WritableFileWriter, and FSWritableFile. When a new FSWritableFile object is created, its io_priority_ can be set dynamically based on the state of the WriteController. In WritableFileWriter, before the call sites of FSWritableFile functions, WritableFileWriter::DecideRateLimiterPriority() determines the rate_limiter_priority. The options (IOOptions) argument of FSWritableFile functions will be updated with the rate_limiter_priority. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9988 Test Plan: Add unit tests. Reviewed By: anand1976 Differential Revision: D36395159 Pulled By: gitbw95 fbshipit-source-id: a7c82fc29759139a1a07ec46c37dbf7e753474cf
3 years ago
writable_file->SetIOPriority(GetRateLimiterPriority());
writable_file->SetWriteLifeTimeHint(write_hint_);
FileTypeSet tmp_set = db_options_.checksum_handoff_file_types;
writable_file->SetPreallocationBlockSize(static_cast<size_t>(
sub_compact->compaction->OutputFilePreallocationSize()));
const auto& listeners =
sub_compact->compaction->immutable_options()->listeners;
outputs.AssignFileWriter(new WritableFileWriter(
std::move(writable_file), fname, fo_copy, db_options_.clock, io_tracer_,
db_options_.stats, listeners, db_options_.file_checksum_gen_factory.get(),
Using existing crc32c checksum in checksum handoff for Manifest and WAL (#8412) Summary: In PR https://github.com/facebook/rocksdb/issues/7523 , checksum handoff is introduced in RocksDB for WAL, Manifest, and SST files. When user enable checksum handoff for a certain type of file, before the data is written to the lower layer storage system, we calculate the checksum (crc32c) of each piece of data and pass the checksum down with the data, such that data verification can be down by the lower layer storage system if it has the capability. However, it cannot cover the whole lifetime of the data in the memory and also it potentially introduces extra checksum calculation overhead. In this PR, we introduce a new interface in WritableFileWriter::Append, which allows the caller be able to pass the data and the checksum (crc32c) together. In this way, WritableFileWriter can directly use the pass-in checksum (crc32c) to generate the checksum of data being passed down to the storage system. It saves the calculation overhead and achieves higher protection coverage. When a new checksum is added with the data, we use Crc32cCombine https://github.com/facebook/rocksdb/issues/8305 to combine the existing checksum and the new checksum. To avoid the segmenting of data by rate-limiter before it is stored, rate-limiter is called enough times to accumulate enough credits for a certain write. This design only support Manifest and WAL which use log_writer in the current stage. Pull Request resolved: https://github.com/facebook/rocksdb/pull/8412 Test Plan: make check, add new testing cases. Reviewed By: anand1976 Differential Revision: D29151545 Pulled By: zhichao-cao fbshipit-source-id: 75e2278c5126cfd58393c67b1efd18dcc7a30772
4 years ago
tmp_set.Contains(FileType::kTableFile), false));
TableBuilderOptions tboptions(
*cfd->ioptions(), *(sub_compact->compaction->mutable_cf_options()),
cfd->internal_comparator(), cfd->int_tbl_prop_collector_factories(),
sub_compact->compaction->output_compression(),
Add more LSM info to FilterBuildingContext (#8246) Summary: Add `num_levels`, `is_bottommost`, and table file creation `reason` to `FilterBuildingContext`, in anticipation of more powerful Bloom-like filter support. To support this, added `is_bottommost` and `reason` to `TableBuilderOptions`, which allowed removing `reason` parameter from `rocksdb::BuildTable`. I attempted to remove `skip_filters` from `TableBuilderOptions`, because filter construction decisions should arise from options, not one-off parameters. I could not completely remove it because the public API for SstFileWriter takes a `skip_filters` parameter, and translating this into an option change would mean awkwardly replacing the table_factory if it is BlockBasedTableFactory with new filter_policy=nullptr option. I marked this public skip_filters option as deprecated because of this oddity. (skip_filters on the read side probably makes sense.) At least `skip_filters` is now largely hidden for users of `TableBuilderOptions` and is no longer used for implementing the optimize_filters_for_hits option. Bringing the logic for that option closer to handling of FilterBuildingContext makes it more obvious that hese two are using the same notion of "bottommost." (Planned: configuration options for Bloom-like filters that generalize `optimize_filters_for_hits`) Recommended follow-up: Try to get away from "bottommost level" naming of things, which is inaccurate (see VersionStorageInfo::RangeMightExistAfterSortedRun), and move to "bottommost run" or just "bottommost." Pull Request resolved: https://github.com/facebook/rocksdb/pull/8246 Test Plan: extended an existing unit test to exercise and check various filter building contexts. Also, existing tests for optimize_filters_for_hits validate some of the "bottommost" handling, which is now closely connected to FilterBuildingContext::is_bottommost through TableBuilderOptions::is_bottommost Reviewed By: mrambacher Differential Revision: D28099346 Pulled By: pdillinger fbshipit-source-id: 2c1072e29c24d4ac404c761a7b7663292372600a
4 years ago
sub_compact->compaction->output_compression_opts(), cfd->GetID(),
cfd->GetName(), sub_compact->compaction->output_level(),
bottommost_level_, TableFileCreationReason::kCompaction,
0 /* oldest_key_time */, current_time, db_id_, db_session_id_,
sub_compact->compaction->max_output_file_size(), file_number);
outputs.NewBuilder(tboptions);
LogFlush(db_options_.info_log);
return s;
}
void CompactionJob::CleanupCompaction() {
for (SubcompactionState& sub_compact : compact_->sub_compact_states) {
sub_compact.Cleanup(table_cache_.get());
}
delete compact_;
compact_ = nullptr;
}
namespace {
void CopyPrefix(const Slice& src, size_t prefix_length, std::string* dst) {
assert(prefix_length > 0);
size_t length = src.size() > prefix_length ? prefix_length : src.size();
dst->assign(src.data(), length);
}
} // namespace
void CompactionJob::UpdateCompactionStats() {
assert(compact_);
Compaction* compaction = compact_->compaction;
compaction_stats_.stats.num_input_files_in_non_output_levels = 0;
compaction_stats_.stats.num_input_files_in_output_level = 0;
for (int input_level = 0;
input_level < static_cast<int>(compaction->num_input_levels());
++input_level) {
if (compaction->level(input_level) != compaction->output_level()) {
UpdateCompactionInputStatsHelper(
&compaction_stats_.stats.num_input_files_in_non_output_levels,
&compaction_stats_.stats.bytes_read_non_output_levels, input_level);
} else {
UpdateCompactionInputStatsHelper(
&compaction_stats_.stats.num_input_files_in_output_level,
&compaction_stats_.stats.bytes_read_output_level, input_level);
}
}
assert(compaction_job_stats_);
compaction_stats_.stats.bytes_read_blob =
compaction_job_stats_->total_blob_bytes_read;
compaction_stats_.stats.num_dropped_records =
compaction_stats_.DroppedRecords();
}
void CompactionJob::UpdateCompactionInputStatsHelper(int* num_files,
uint64_t* bytes_read,
int input_level) {
const Compaction* compaction = compact_->compaction;
auto num_input_files = compaction->num_input_files(input_level);
*num_files += static_cast<int>(num_input_files);
for (size_t i = 0; i < num_input_files; ++i) {
const auto* file_meta = compaction->input(input_level, i);
*bytes_read += file_meta->fd.GetFileSize();
compaction_stats_.stats.num_input_records +=
static_cast<uint64_t>(file_meta->num_entries);
}
}
void CompactionJob::UpdateCompactionJobStats(
const InternalStats::CompactionStats& stats) const {
compaction_job_stats_->elapsed_micros = stats.micros;
// input information
compaction_job_stats_->total_input_bytes =
stats.bytes_read_non_output_levels + stats.bytes_read_output_level;
compaction_job_stats_->num_input_records = stats.num_input_records;
compaction_job_stats_->num_input_files =
stats.num_input_files_in_non_output_levels +
stats.num_input_files_in_output_level;
compaction_job_stats_->num_input_files_at_output_level =
stats.num_input_files_in_output_level;
// output information
compaction_job_stats_->total_output_bytes = stats.bytes_written;
compaction_job_stats_->total_output_bytes_blob = stats.bytes_written_blob;
compaction_job_stats_->num_output_records = stats.num_output_records;
compaction_job_stats_->num_output_files = stats.num_output_files;
compaction_job_stats_->num_output_files_blob = stats.num_output_files_blob;
if (stats.num_output_files > 0) {
CopyPrefix(compact_->SmallestUserKey(),
CompactionJobStats::kMaxPrefixLength,
&compaction_job_stats_->smallest_output_key_prefix);
CopyPrefix(compact_->LargestUserKey(), CompactionJobStats::kMaxPrefixLength,
&compaction_job_stats_->largest_output_key_prefix);
}
}
void CompactionJob::LogCompaction() {
Compaction* compaction = compact_->compaction;
ColumnFamilyData* cfd = compaction->column_family_data();
// Let's check if anything will get logged. Don't prepare all the info if
// we're not logging
if (db_options_.info_log_level <= InfoLogLevel::INFO_LEVEL) {
Compaction::InputLevelSummaryBuffer inputs_summary;
ROCKS_LOG_INFO(
db_options_.info_log, "[%s] [JOB %d] Compacting %s, score %.2f",
cfd->GetName().c_str(), job_id_,
compaction->InputLevelSummary(&inputs_summary), compaction->score());
char scratch[2345];
compaction->Summary(scratch, sizeof(scratch));
ROCKS_LOG_INFO(db_options_.info_log, "[%s]: Compaction start summary: %s\n",
cfd->GetName().c_str(), scratch);
// build event logger report
auto stream = event_logger_->Log();
stream << "job" << job_id_ << "event"
<< "compaction_started"
<< "compaction_reason"
<< GetCompactionReasonString(compaction->compaction_reason());
for (size_t i = 0; i < compaction->num_input_levels(); ++i) {
stream << ("files_L" + std::to_string(compaction->level(i)));
stream.StartArray();
for (auto f : *compaction->inputs(i)) {
stream << f->fd.GetNumber();
}
stream.EndArray();
}
stream << "score" << compaction->score() << "input_data_size"
<< compaction->CalculateTotalInputSize() << "oldest_snapshot_seqno"
<< (existing_snapshots_.empty()
? int64_t{-1} // Use -1 for "none"
: static_cast<int64_t>(existing_snapshots_[0]));
if (compaction->SupportsPerKeyPlacement()) {
stream << "preclude_last_level_min_seqno"
<< preclude_last_level_min_seqno_;
stream << "penultimate_output_level" << compaction->GetPenultimateLevel();
stream << "penultimate_output_range"
<< GetCompactionPenultimateOutputRangeTypeString(
compaction->GetPenultimateOutputRangeType());
if (compaction->GetPenultimateOutputRangeType() ==
Compaction::PenultimateOutputRangeType::kDisabled) {
ROCKS_LOG_WARN(
db_options_.info_log,
"[%s] [JOB %d] Penultimate level output is disabled, likely "
"because of the range conflict in the penultimate level",
cfd->GetName().c_str(), job_id_);
}
}
}
}
std::string CompactionJob::GetTableFileName(uint64_t file_number) {
return TableFileName(compact_->compaction->immutable_options()->cf_paths,
file_number, compact_->compaction->output_path_id());
}
Set Write rate limiter priority dynamically and pass it to FS (#9988) Summary: ### Context: Background compactions and flush generate large reads and writes, and can be long running, especially for universal compaction. In some cases, this can impact foreground reads and writes by users. From the RocksDB perspective, there can be two kinds of rate limiters, the internal (native) one and the external one. - The internal (native) rate limiter is introduced in [the wiki](https://github.com/facebook/rocksdb/wiki/Rate-Limiter). Currently, only IO_LOW and IO_HIGH are used and they are set statically. - For the external rate limiter, in FSWritableFile functions, IOOptions is open for end users to set and get rate_limiter_priority for their own rate limiter. Currently, RocksDB doesn’t pass the rate_limiter_priority through IOOptions to the file system. ### Solution During the User Read, Flush write, Compaction read/write, the WriteController is used to determine whether DB writes are stalled or slowed down. The rate limiter priority (Env::IOPriority) can be determined accordingly. We decided to always pass the priority in IOOptions. What the file system does with it should be a contract between the user and the file system. We would like to set the rate limiter priority at file level, since the Flush/Compaction job level may be too coarse with multiple files and block IO level is too granular. **This PR is for the Write path.** The **Write:** dynamic priority for different state are listed as follows: | State | Normal | Delayed | Stalled | | ----- | ------ | ------- | ------- | | Flush | IO_HIGH | IO_USER | IO_USER | | Compaction | IO_LOW | IO_USER | IO_USER | Flush and Compaction writes share the same call path through BlockBaseTableWriter, WritableFileWriter, and FSWritableFile. When a new FSWritableFile object is created, its io_priority_ can be set dynamically based on the state of the WriteController. In WritableFileWriter, before the call sites of FSWritableFile functions, WritableFileWriter::DecideRateLimiterPriority() determines the rate_limiter_priority. The options (IOOptions) argument of FSWritableFile functions will be updated with the rate_limiter_priority. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9988 Test Plan: Add unit tests. Reviewed By: anand1976 Differential Revision: D36395159 Pulled By: gitbw95 fbshipit-source-id: a7c82fc29759139a1a07ec46c37dbf7e753474cf
3 years ago
Env::IOPriority CompactionJob::GetRateLimiterPriority() {
if (versions_ && versions_->GetColumnFamilySet() &&
versions_->GetColumnFamilySet()->write_controller()) {
WriteController* write_controller =
versions_->GetColumnFamilySet()->write_controller();
if (write_controller->NeedsDelay() || write_controller->IsStopped()) {
return Env::IO_USER;
}
}
return Env::IO_LOW;
}
} // namespace ROCKSDB_NAMESPACE